{

$-

4’
23::
.

  
  

'
‘

'1‘

$

'3‘

l.-
.
4‘
.4

a... .
.....p a...”

-1

Wu

:7}; tab

’:'::T A
-.._.

   
  
 

 

.‘.LJ.'«

  

u-.-—. -
4‘21:
.- ..'.,-.

.1w4—‘

   
 
  
 
 
  

 

L
!‘

 

 

 

 
 

\vu
PM
2.5!:

 
  

    

   
  
 
 
 
 
 
   

Jay
4 fig.

0
‘ 411‘ 1
{”3" $5}

 

  
  
    
 

 

 

f’iem

i
mg;

+3 ’1. {7,1 ‘E"

2....
dazzlp‘

 
  
  

i
. I' ‘
“M?

 
 
   
 
  
 
   

§.

{'4
5“ (g!

  
   

  
  

      

1:2;- 1.1.
if?! MM n I ism“; “if ' 3.

“'6: :ytawi, gem?
rt“ ‘ u ‘ ‘ ‘1}; .I“

'23???
KN; 4:
n; I

4.—

11:1.

"07-

        
     
 

v

     

3

. 1". flan“?

u‘ql,‘ 31‘1”? u-g; fl?" 9333'
v '::""» ‘0 ."S .'.'
”333' €5"".’!¢Ms!fl~ 311113 ,

a; .
. :J :, A: ”$311.;
3? 21:“???3 333v! . w:
" -: 7:. ‘

"311 ~ h

4: ~ _1

‘3“: i ’é‘gtz‘I-I‘
We!" 1 ' -'
'i‘

    
  

   
  

       

 

 
  
 

 
 

.‘J

   
  

 
 

 
   

 
 

- .gf'.t

- 1'2"

 

   
  
  
  
  
 
 

     

   

   
 
   

      
    

gzi'wf

3:25. .‘ :32...

j :' n .1 i".-

I 2' Wm .
' ":5 '- ‘ 1:53.451:
3;} ,-; ' L . ‘ ' 11““:
J - It", ‘ .' ‘. ‘F;

t " . a,”

331m

‘ 4’s?

.: :3

. . .. > :9
. W “t.“ :1 . ,-
.;. hi I .ri.
if; A}
. z f

.‘L‘

   

'33-
I..-‘
x‘
'.

  

YHES‘S

RSITY UBRARIES

“W M .mmmm Mixing H\ \\|

31293307

 

 

 

 

 

 

m N

This is to certify that the
dissertation entitled

NUCLEAR EXPORT OF A PRE-mRNA SPLICING FACTOR
GALECTIN-3

presented by

YEOU-GUANG TSAY

has been accepted towards fulfillment
of the requirements for

 

Eh. 12. degree in WM
1 , Lie—1
U Major professor )

Date X'22'77

.MSU is an Affirmative Action/Equal Opportunity Institution 0- 12771

 

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN BOX
to remove this checkout from your record.
TO AVOID FINES return on or before date due.

 

MTE DUE MTE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1/93 campus-p.14

 

NUCLEAR EXPORT OF A PRE-mRNA SPLICING FACTOR, GALECTIN-3
By

Yeou-Guang Tsay

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Biochemistry

1 997

ABSTRACT

NUCLEAR EXPORT OF A PRE-mRNA SPLICING FACTOR, GALECTIN-3

By

Yeou-Guang Tsay

Previous immunofluorescence and subcellular fractionation studies have
documented that galectin—3 (Mr ~ 30,000) distributes differentially between the nucleus
and the cytoplasm, depending on the proliferative state of the cells under analysis. Thus,
it appears that controlled nuclear versus cytoplasmic distribution of galectin-3 may be one
of the mechanisms regulating its function. The goal of this thesis was to study
parameters that play a role in determining this distribution; one such parameter turned out

to be the export of the protein from the nucleus to the cytoplasm.

Using digitonin-penneabilized 3T3 cells, we provide evidence that galectin-3 is
rapidly and selectively exported from the nucleus. Although both phosphorylated and
nonphosphorylated isoforms of galectin-3 are found in the nuclear fiaction, only
phosphorylated galectin-3 is identified in the exported fraction, implying that
phosphorylation is important forthe nuclear export of the protein. The rate of galectin-3
export is decreased by cold temperature and by the addition of either wheat germ
agglutinin or vanny cations (V 0"). More strikingly, galectin-3 export can be inhibited,

at least partially, by the simultaneous addition of a peptide bearing a nuclear export signal

plus a dinucleotide analog of the cap structure found at the 5’-end of mRNAs. These
results suggest that galectin-3 may be exported in association with ribonucleoprotein
complexes containing monomethylated cap structure as well as polypeptides containing

nuclear export sequences.

The transported fraction of the nuclear export assay was analyzed in terms of the
polypeptide and RNA components. Gel filtration of the exported nuclear material and
analysis for galectin-3 showed that the lectin can be found in at least two sets of high
molecular weight complexes (~650 RD and ~60 kD). In the presence of the saccharide
ligand, lactose, both of these complexes are disrupted and galectin-3 chromatographs to a
position corresponding to ~30 kD polypeptide. The polypeptide components of the high
molecular weight complexes containing galectin-3 are specifically revealed by affinity
adsorption on a lactose-agarose column, specific elution by lactose, and gel

electrophoresis. These polypeptides are not bound to a control cellobiose column.

The transported fraction of the nuclear export assay also contains RNA. In the
low molecular weight range, the RNA species include tRNA (~80 nucleotides) as well as
RNAs of ~100, 300, and 650 nucleotides. High molecular weight RNAs, ranging from
~1 kb to 5 kb, include poly (A)+ mRNA as revealed by hybridization with an oligo(dT)
probe. Compounds containing vanadyl cations, which inhibit the export of galectin-3,
also inhibit the export of RNAs. All of these results are consistent with the notion that
galectin-3 is associated with a ribonucleoprotein complex during its export fiom the

nucleus.

Copyright by
Yeou-Guang Tsay
1 997

To my parents and Huifen

ACKNOWLEDGMENTS

I would like to express my gratitude to my mentor, Dr. John Wang, who has
provided me every available opportunity that can help me to become a better scientist. I
enjoyed all of the discussions that we had in these years, which have broadened my view
over the field greatly.

I would like to thank my guidance committee, Drs. Don Jump, Ron Patterson, El
Schindler, and Bill Smith for their criticism and advice, which has been a major part of
this work.

I thank former and present members of Wang and Schindler’s labs for their
friendship and encouragement, especially Dr. Eric Arnoy, Sharon Grabski, Nancy Lin,
Mark Kadrofske, Patty Voss and Dr. Anandita Vyakarnam. They have provided generous
help to myself as well as my family. Special thanks will be given to Patty and Nancy,
who have been helping me to complete my dissertation.

Finally, I feel very grateful to my wife. She has accompanied me through the ups
and downs in these years. With her love and support, life is much more pleasant and

colorful.

TABLE OF CONTENTS

TABLE OF CONTENTS ....................................................................... vii
LIST OF TABLES ................................................................................ ix
LIST OF FIGURES ............................................................................... x
LIST OF ABBREVIATIONS ................................................................. xiii
CHAPTER I: LITERATURE REVIEW ...................................................... 1

INTRODCUTION TO GALECTINS ................................................. 1

INTRODUCTION TO NUCLEAR TRANSPORT ................................ 22

CHAPTER II: NUCLEAR EXPORT OF GLAECTIN-3 IN MOUSE 3T3
FIBROBLASTS: PARAMETERS OF THE TRANSPORT IN

DIGITONIN-PERMEABILIZED CELLS ................................. 44
SUMMARY ...................................................... 45
INTRODUCTION ....................................................................... 46
EXPERIMENTAL PROCEDURES ................................................... 50
RESULTS ................................................................................ 58
DISCUSSION ............................................................................ 74

CHAPTER III: NUCLEAR EXPORT OF GALECTIN-3 IN MOUSE 3T3
FIBROBLASTS: CHARACTERIZATION OF THE EXPORTED

COMPONENTS ................................................................. 80
SUMMARY ............................................................................... 81

INTRODUCTION ....................................................................... 82

vii

EXPERIMENTAL PROCEDURES .................................................. 84

RESULTS ................................................................................ 89
DISCUSSION .......................................................................... 106
CONCLUSIONS .............................................................................. 1 10
LIST OF REFERENCES ..................................................................... 112

viii

CHAPTER I

4.

LIST OF TABLES

SUMMARY OF GALECTINS ................................................ 3
VERTEBRATE NUCLEAR PORE PROTEINS .......................... 25
SIGNALS AND RECEPTORS INVOLVED IN NUCLEAR

IMPORT ......................................................................... 29

SIGNALS AND RECEPTORS INVOVLED IN NUCLEAR EXPORT
.................................................................................... 38

i)

CHAPTER I

CHAPTER II

1.

LIST OF FIGURES

A PHYLOGENETIC TREE FOR CROWN EUKARYOTES ............ 2

INTERACTION OF BOVINE GALECTIN-l WITH N-
ACETYLLACTOSAMINE .................................................... 6

SCHEMATIC REPRESENTATION OF THE INVOVLEMENT OF
SMALL RIBONUCLEOPROTEIN PARTICLES (snRNP) AND NON-
snRNP PROTEINS IN PRE-mRNA SPLICING .......................... 17

ROLES OF GALECTIN CRD IN PRE-mRNA SPLICING ............. 20

SCHEMATIC REPRESENTATION OF A CONCENSUS MODEL OF
THE MEMBRANE-BOUND NUCLEAR PORE COMPLEX .......... 23

DIAGRAM SUMMARIZING LOCALIZATIONS OF
CHARACTERIZED NUCLEOPORINS WITHIN THE NUCLEAR
PORE COMPLEX ............................................................. 27

NUCLEAR IMPORT OF NLS-CONTAININ G PROTEINS HAS TWO
DISTINCT STEPS: DOCKING AND TRANSLOCATION ............ 31

(A) A MODEL ILLUSTRATING RAN GTPASE CYCLE. (B)
COORDINATION OF PROTEIN IMPORT BY RAN GTPASE
CYCLE ........................................................................... 33

SCHEMATIC ILLUSTRATION OF THE DIGITONIN-
PERMEABILIZATION CELL SYSTEM FOR THE ASSAY OF
NUCLEAR EXPORT ......................................................... 53

COMPARISON OF THE IMMUNOFLUORESCENCE STAINING
FOR GALECTIN-3 IN 3T3 CELLS BEFORE AND AFTER
DIFFERENTIAL PERMEABILIZATION ................................. 59

CHAPTER III

1.

WESTERN BLOTTING ANALYSIS FOR GALECTIN-3 IN THE
NUCLEAR FRACTION (NF) AND THE TRANSPORTED
FRACTION (TF) BEFORE AND AFTER A 30-MIN EXPORT
ASSAY ........................................................................... 61

TEMPERATURE DEPENDENCE OF THE KINETICS OF
GALECTIN-3 EXPORT IN THE DIGITONIN PERMEABILIZED
CELL ASSAY .................................................................. 63

TWO-DIMENSIONAL GEL ELECTROPHORETIC ANALYSIS FOR
GALECTIN-3 IN NF BEFORE THE EXPORT ASSAY (A) AND IN
THE TF AFTER AN 8-MIN EXPORT ASSAY (B) ....................... 66

COMPARISON OF SEVERAL NUCLEAR PROTEINS IN TERMS
OF THEIR EXPORT OR RETENTION IN THE DIGITONIN-
PERMEABILIZED CELL ASSAY ......................................... 68

THE EFFECT OF WHEAT GERM AGGLUTININ (WGA) ON THE
EXPORT OF GALECTIN-3 AND RAN IN THE DIGITONIN-
PERMEABILIZED CELL ASSAY ......................................... 70

THE EFFECTS OF S’-CAP STRUCTURE NUCLEOTIDE ANALOGS
AND OF A PEPTIDE BEARING A NUCLEAR EXPORT SIGNAL
(NES) ON THE EXPORT OF GALECTIN-3 AND RAN IN THE
DIGITONIN-PERMEABILIZED CELL ASSAY ........................ 72

SCHEMATIC DIAGRAM ILLUSTRATING THE ASSOCIATION OF
GALECTIN-3 WITH A RIBONUCLEOPROTEIN COMPLEX IN TF
OF THE EXPORT ASSAY ................................................... 73

GELFILTRATION OF THE EXPORTED NUCLEAR COMPONENTS
IN TF AND ANALYSIS FOR GALECTIN-l AND -3 ................... 91

THE EFFECT OF LAC ON THE GEL FILTRATION PROFILE OF
GALECTIN-l AND GALECTIN-3 IN TF ................................. 93

AFFINITY ADSORPTION OF THE POLYPEPTIDE COMPONENTS
IN TF ON LACTOSE AGAROSE AND CELLOBIOSE-AGAROSE
..................................................................................... 95

THE EFFECT OF VANADYL CATIONS (V 02+) ON THE NUCLEAR
EXPORT OF GALECTIN-3. ASSAYED BY IMMUNOBLOT'TING
NF ................................................................................. 99

THE EFFECT OF VANAYL CATIONS (V 02*) ON THE NUCLEAR
EXPORT OF GALECTIN-3 BY IMMUNOFLUORESCENCE
ANALYSIS OF NF ........................................................... 100

THE EFFECTS OF VANADYL CATIONS (V 02+) ON THE
NUCLEAR EXPORT OF RAN, ASSAYED BY
IMMUNOFLUORESCENCE ANALYSIS OF NF ...................... 102

7. GEL ELECTROPHORETIC ANALYSIS OF THE RNA

COMPONENTS IN TF WHEN EXPORT ASSAY IS CARRIED OUT
IN THE ABSENCE AND PRESENCE OF VANDYL SULFATE. ....103

xii

ATP:

BR:

CBC:

CBP:

CRD:

DNA:

ER:

FGF:

FITC:

GDP:

GlcNAc:

GTP:

HIV:

hnRNP:

Lac:

WCK:

MIP:

mRNA:

NEPHGE:

LIST OF ABBREVIATIONS

adenosine triphosphate

Balbiani ring

cap binding complex

cap binding protein

carbohydrate recognition domain
deoxyribonucleic acid
endoplasmic reticulum

fibroblast growth factor
fluorescein isothiocyanate
guanosine diphosphate
N-acetylglucosamine

guanosin tn'phosphate

human immunodeficiency virus
heterogeneous ribonucleoprotein complex
lactose

Madin-Darby canine kidney

M9 interacting protein
messenger ribonucleic acid

nonequilibrium pH gradient electrophoresis

xiii

NES:

NTF:

PAGE:

PKI:

RanGAP:

RIP:

SAP:

SDS:

snRNP:

SV40:

T-TBS:

WGA:

nuclear export signal

nuclear fraction

nuclear localization signal
nuclear pore complex

nuclear transport factor
polyacrylamide gel electrophoresis
protein kinase A inhibitor

Rev activation domain

Ran GTPase activating protein
Rev interacting protein
ribonucleoprotein
spliceosome-associated protein
sodium dodecyl sulfate

small nuclear ribonucleoprotein
simian virus 40

transported buffer

transported fraction
thimerosol-Tris buffered saline
vanadyl ribonucleoside complex

wheat germ agglutinin

xiv

CHAPTER 1

LITERATURE REVIEW
INTRODUCTION TO GALECT INS

Galectins are a family of proteins defined by their affinity toward
galactose/lactose and sequence similarity at the characteristic carbohydrate binding
domain (Barondes er al., 1994; Kasai and Hirabayshi, 1996; Leffler, 1997). Galectins
were previously found in species in the animal kingdom, including sponges, nematodes,
fish, amphibian, birds and mammals. Hence, galectins were considered as a group Of
animal lectins for years (Kasai and Hirabayashi, 1996). However, the discovery of a
galectin in inky cap mushroom, Coprinus cinereus, challenged this concept (Cooper et
al., 1997). This finding not only confirms that galectins can be outside the animal
kingdom, but also implies that a common ancestor for animals and fungi may contain
galectin-like polypeptides (Figure 1).

The evolution of galectins has been an interesting topic. For instance, they have
been expanded into a large group of homologs in mammals (Table I), in which at least ten
galectins have been identified (Table 1). On the contrary, it is surprising to see that
galectins seem to be completely abandoned in some descendants, like yeasts and insects

(Cooper et al., 1997) .

 

 

 

Plants

 

 

Prolifera
(Sponges)

Anthropoda
(Insects)

Nematoda
(Nematodes)

Chorda
(Birds, Amphibians,
Fish, Mammals)

Basidiomycota
(mushroom)

Ascomycota
(Yeasts)

Figure l. A phylogenetic tree for crown eukaryotes. Galectins have been found in

animals, like sponges, nematodes, and vertebrates, as well as a fungus, inky cap

mushroom.

Table 1. Summary of Galectins

 

 

Designations MW Structur Tertiary Source Organisms Tissue/Cell Distribution
(kDa) :11 Type Structure
Mammals
Galectin-l 14.5 Proto Dimer Human, rat, mouse, Muscle, heart, lung,
hamster, monkey, ox, placenta, brain, spleen,
pig liver, lymph node, thymus,
prostate, colon
Galectin-Z 14.5 Proto Dimer Human, mouse Small intestine
Galectin-3 29-35 Chimera Monomer Human, rat, mouse, Macrophage, colon,
dog, hamster leukemia cell, fibroblasts
Galectin-4 36 Tandem Monomer Human, rat, pig, Alimentary tract epithelium
repeat mouse
Galectin-S 17-18 Proto Monomer Rat Erythrocyte
Galectin-6 34 Tandem Monomer Mouse Castro-intestine
repeat
Galectin-7 14.5 Proto ? Human, rat Skin
Galectin-8 34 Tandem Monomer Human, rat Liver, lung, kidney
repeat
Galectin-9 35 Tandem ? . Human, rat, mouse Kidney, thymus, Hodgkin's
repeat lymphoma
Galectin-IO 17 Proto Dimer Human
Birds
Chick 14K 14 Proto Monomer Chick Skin, intestine, etc.
Chick 16K 16 Proto Dimer Chick Muscle, liver, etc.
Chick 30K 30 Chimera ? Chick Chondrocyte
Amphibians
Xenopus 16K 16 Proto Dimer Xenopus Iaevis Skin
Bufo 15K 14.5 Proto Dimer B. Arenarum Oocyte
Fish
Electrolectin l6 Proto Dimer Electric eel Electric organ
Congerin l6 Proto Dimer Conger eel Skin mucus
Nematodes
Nematode 32K 32 Tandem Monomer C. elegans Cuticle, pharynx
repeat ,
Nematode 16K 16 Proto Dimer C. elegans ?
OvaGalBP 32 Tandem 7 0. volvulus ?
repeat
sponges
Gth1/2 13-18 Proto Dimer G. cydonium Plasma membrane
Fungi
Cgl-I/II 15.5/ Proto Dimer C. cinereus Fruiting body

17

4

The galectin structure also went through some changes. Chimera-type galectins
are only found in vertebrates but not in primitive animals, implying that they are evolved
later than proto- and tandem-repeat type galectins (Table 1). In addition, it seems that
activities other than carbohydrate binding has been introduced into the galectin structure.
For instance, galectin-10, the Charcot-Leyden crystal protein, was reported to have a
lysophospholipase activity (Weller et al., 1982; Ackerman et al., 1993).

In the following section, I will focus on the structural characteristics, intracellular
sorting, and biological functions of galectins. Since galectin-l and galectin-3 are the
most studied galectins, I will discuss properties of the whole family using them as

examples.

STRUCTURAL CHARACTERISTICS

Galectins are categorized into proto-, chimera-, and tandem-repeat types
according to their primary structure. According to their molecular architecture, all
galectins can be grouped into the following three types: proto-, tandem-repeat and
chimera types (Hirabayashi et al., 1992) (Table 1). Prototype, represented by mammalian
galectin-1 is composed of only a single carbohydrate recognition domain (CRD) motif.
Tandem-repeat type galectins, exemplified by galectin-4, consist of two homologous
CRD domains. Chimera-type galectin, represented by galectin-3 is made of a CRD and a
nOn-lectin domain.

Carbohydrate recognition domain is the signature motif for galectin family.
The galectin family has a signature structural motif, carbohydrate recognition domain,

which confers saccharide-binding specificity. This domain consists of about 135 amino

5

acids that fold into a B-sandwich with 5- and 6-stranded B-sheets as shown by X-ray
diffraction studies on galectin-1 and galectin-2 (Lobsanov et al., 1993; Liao et al., 1994).
All of the amino acid residues important for binding with the bound N-acetyllactosamine
ligand are found on a continuous stretch of the polypeptide chain, which makes up four
antiparallel B-strands and is designated as carbohydrate-binding cassette (Gitt et al.,
1992; Lobsanov and Rini, 1997). Galectins forming homodimers, like galectin-l and
galectin-2, may contain another structural motif, dimer interface. This motif is formed by
two B—strands at the end away from the carbohydrate-binding cassette .

Carbohydrate-binding cassette is conserved among galectin family. Amino
acid residues that are involved in carbohydrate binding are highly conserved among
galectins (Lobsanov et al., 1993; Liao et al., 1994). Analysis of the genomic DNA
structure of a number of galectins has also shown that they are all included on a single
DNA exon (Barondes et a1. , 1994; Lobsanov and Rini, 1997).

There are eight major amino acids in CRDs that directly interact with a
disaccharide ligand, like N-acetyllactosamine (Lobsanov et al., 1993; Liao et al., 1994;
Ahmed and Vasta, 1994). Among these amino acids, five residues are invariably
conserved in all galectins. They primarily interact with the lactose moiety of the ligand
(Figure 2). On the other hand, a lot of variations occur at the other three amino acids.
They primarily interact with the N-acetylglucosamine moiety of the ligand.

Ahmed and Vasta (1994) found that complete identity with galectin-l CRD at the
latter three residues is found in a number of the galectins, whereas substitution and

deletion occur at these positions in other galectins. Based on identity of these three

 

Figure 2. Interaction of bovine galectin-l with N-acetyllactosamine. Amino acids
boxed by lighter shade are completely conserved in all galectins except mammalian
galectin-10. Amino acids highlighted by darker shade have greater variations among
galectin family. (Modified from Ahmed and Vasta, 1994)

7

amino acids with the galectin-1 CRD, carbohydrate recognition domains of galectins can
be classified as two types: type I (conserved) and type II (variable) CRDs (Ahmed and
Vasta, 1994). So far, only chicken 14-kDa, chicken 16 kDa, and Bufo 15-kDa galectins
belong to the type I category. Type I CRDS have very similar fine carbohydrate-binding
specificity, presumably due to the extensive identities at the carbohydrate-binding
cassette. On the other hand, type II CRDs have very different carbohydrate recognition
properties, reflecting the variations in the cassette motif.

The carbohydrate-binding cassette in galectin-10 is unusually varied at these
positions, compared with other galectins (Ackerman et al., 1993). Different amino acids
are found at the positions that are completely conserved in other galectins. This is
perhaps the structural basis for its weaker sugar binding activity. Probably, in order to
accommodate a lysophospholipase activity as well as lectin activity, some compromise
had to be made on the primary structure.

Galectin-l/galectin-2 dimer interface mediates the homodimerization of
galectin subunits. The dimer interface for galectin-1 and galectin-2 is mainly made up
Of hydrophobic amino acid residues, which form nonpolar surfaces (Lobsanov et al.,
1993; Liao et al., 1994). This interface mediates the formation of a 2-fold symmetric
anti-parallel dimers seen in galectin-1 and galectin-2. Recent x-ray crystallography of
galectin-3 CRD (Kanigsberg et al. 1997) and N-tenninal CRD of galectin-4 (Kayden et
al., 1997) have shown that the nonpolar surfaces are essentially eliminated by the
presence of hydrophilic residues. Lobsanov and Rini (1997) argued that the disruption of
nonpolar surfaces prevents galectin-3 or galectin-4 from forming homodimer complex.

However, whether CRD of galectin-3 or galectin-4 forms dimers is still controversial.

8

For instance, when isolated C-domain of galectin-3 was analyzed by gel filtration
chromatography, dimer formation was found in a significant fraction of galectin-3 CRD
(Wang et al., 1993). Therefore, whether this dimer interface is exclusively found for
prototype galectins like galectin-l and galectin-Z-needs further examination.

Galectin-3, a chimera-type galectin, contains a proline-glycine-rich domain.
Galectin-3 contains a proline-glycine-rich domain at the NHz-terminal end (J ia and Wang,
1988), in addition to a CRD. This N-domain is composed mainly of sequence repeats,
each consisting Of nine amino acids (PGAYPGXXX) in mammalian homologues (J ia and
Wang, 1988) and eight amino acids (PGPYPGGP) in the chicken counterpart
(Nurminskaya and Linsenmayer, 1996). So far, the role(s) of this domain is not clear,
although similar features have also been seen in some other proteins, like a spliceosome-
associated protein (SAP) SAP62 (Bennet and Reed, 1993) and an autoantigen annexin XI
(Misaki et al., 1994).

SAP62 is a defined component of spliceosome complexes E, A and B (Bennett et
al., 1992; Hong et al., 1997). This splicing factor contains a C-domain of 22 proline-
glycine-rich heptapeptide repeats with the consensus sequence PGVHPPA (Bennett and
Reed, 1993) besides the Zn fmger-containing N-domain. It has been proposed that the P-,
G-rich domain may be involved in protein-protein interaction (Bennett and Reed, 1993).
Nevertheless, it will be interesting to see how it is related to galectin-3 N-domain as both
SAP62 and galectin-3 are involved in pre-mRNA splicing.

Annexin XI, on the other hand, was identified as a calcyclin-associated protein
(Tokumitsu et al., 1992), which is recognized by sera of patients with various

autoimmune diseases (Misaki et al. , 1994). Structurally, it contains a proline-glycine-rich

9

N-terrninal tail domain as well as a core domain that was conserved among annexin
family. The N-terrninal tail has been shown to be responsible for its nuclear localization
(Mizutani et al., 1995). Interestingly, annexin XI, a 56-KDa protein also shows a dual
(nuclear and cytoplasmic) localization (Misaki et al., 1994), which raises the possibility

that galectin-3 N-domain may also have the same intracellular targeting capacity.

INTRACELLULAR SORTING

Subcellular localization of galectins has been extensively studied for galectin-1
and galectin-3. These two galectins, like a large number of other proteins (Smalheiser,
1996), have been found to reside in multiple cellular compartments. Protein targeting
between different compartments is usually via signal-mediated pathways. It is always
interesting to see how different signals are incorporated into the protein structure, and
how a particular signal is selected over others. Localization studies on galectins
definitely may help answer these questions, considering their relatively simple primary
structures. In this section, I will concentrate on the pathways mediating the extracelluar
secretion as well as nucleocytoplasmic distribution.

Galectin-l and galectin-3 are secreted out of cells via the pathway(s) distinct
from the classical pathway. Most extracellular proteins leave the cells using the
classical secretory pathway, which requires a signal peptide on the substrate proteins
(Rothman and Orci, 1992). The secreted proteins enter the ER, and move to the cell
surface via the Golgi apparatus. Both galectin-1 and galectin-3 lack a classical secretory

signal (Muesch et al., 1990) and are secreted through so-called non-classical secretory

10

pathway(s). Examples of proteins utilizing nonclassical pathways include interleukin-113
(Siders and Mizel, 1995), FGF-l (Tarantini et al., 1995), and FGF-2 (Piotrowicz et al.,
1997). Both galectin-l and galectin-3 have been used as a model system to explore these
pathways.

Galectin-l appears to be externalized from cells through extracellular
vesicles derived from evaginations of plasma membrane. Galectin-l has been found
to be secreted at a higher rate in differentiated muscles cells than in myoblasts (Cooper
and Barondes, 1990). Immunofluorescent localization studies showed galectin-l is
primarily cytoplasmic in myoblasts, but cytoplasmic staining is largely lost in myotubes,
the well-differentiated muscle cells. Examination of differentiating myoblasts, the
intermediate between two stages, showed that galectin-1 had a patchy distribution at this
transition. These patches were in the outermost layer of cytoplasm and appeared to
correspond to the evaginations in plasma membranes. Cooper and Barondes (1990)
hypothesized that these evaginations may later mature into extracellular vesicles and
serve as the secretory vehicles for galectin-1.

Yeast cells can also secret galectin-l , when they are transformed with a galectin-
l-expressing vector (Cleves et al., 1996). This Observation indicates that a pathway for
secreting galectin-1 is conserved between budding yeast and mammalian cells. This
secretion does not require any components in the classical secretory pathway nor the
yeast multidrug resistance-like protein Ste6p, an a factor transporter. A genetic screen
was designed to identify gene products which are involved in secretion. Two gene

products, NCEl and NCE2, are able to enhance galectin-l secretion when they are

ll

overexpressed. (Cleves et al., 1996). Furthermore, nce2 null mutants showed a great
reduction in galectin-1 secretion, establishing the pivotal role of NCE2 in nonclassical
pathway for galectin-1. The primary structure of NCE2 may allow the polypeptide to
fold into a membrane protein. Whether this protein is related to the secretion mechanism
seen in muscle cells awaits more documentation.

Galectin-3 is selectively secreted from apical surface, but not from the
basolateral membrane, of Madin-Darby canine kidney (MDCK) cells via a
nonclassical pathway. Plasma membranes of polarized epithelia cells have two
domains, the apical and basolateral, which are separated from each other by tight
junctions (Sirnons and Wandinger-Ness, 1990; Mostov et al., 1992). Distinct sets of
proteins are targeted to each domain. Galectin-3 in MDCK cells is specifically secreted
from the apical surface through a nonclassical pathway (Lindstedt et al., 1993). This
conclusion is supported by the observation that this pathway was not inhibited by
brefeldin A or monensin, drugs that inhibit the ER-Golgi pathway. On the other hand,
the secretion is enhanced by calcium ionophore A23187 and by heat shock at 42°C.
Lowered temperature as well as addition of nocodazole blocks its secretion. All of these
distinguish this pathway from the classical pathway.

Galectin-3 distributes between cytoplasm and the nucleus as a function of
proliferating capacity. In mouse 3T3 fibroblasts, galectin-3 was found both in
cytoplasm and in the nucleus. (Moutsatsos et al., 1986). The distribution between these
two compartments correlates with proliferative states of cells (Moutsatsos et al., 1987).

While in serum-starved cells, galectin-3 is primarily cytoplasmic, it becomes

12

predominantly nuclear as cells are reactivated by serum addition. Similar results can be
seen in dense cultures versus sparse cultures. Sulicellular fractionation studies showed
that galectin-3 in nuclear fraction has two isoforrns, a pI 8.2 phosphorylated form and a pl
8.7 dephosphorylated form. However, only phosphorylated isofonn is seen in cytosolic
fraction. These results suggest that phosphorylation/dephosphorylation is involved in
distribution between two compartments.

Similar results have also been shown on human primary fibroblasts. These cells
are susceptible to senescent change (Hamann et al., 1991; Cowles et al., 1989). In young
cells, galectin-3 is mainly in the nucleus. As cells become senescent, galectin-3 is
exclusively found in cytoplasm.

Altered intracellular distribution has also been seen in neoplastic cells (Lotz et al.,
1993). Paradoxically, galectin-3 is exclusively cytoplasmic in colon carcinoma cells, in
contrast to its predominant nuclear localization in well-differentiated epithelial cells.

Nuclear galectin-3 is associated with the nuclear matrix defined by detergent
and high-salt extraction. In immunofluorescence studies, galectin-3 was found
diffusely distributed throughout the nucleoplasm excluding nucleoli. Even after detergent
and high-salt extraction, it still associates with nuclear structures that contain splicing
factors SC35 and Sm proteins (Wang et al., 1995a; Hubert et al., 1995; Vyakarnam et al.,
1997). The association is dependent on RNA-containing structures. Consistent with
these results, galectin-3 has been found in purified nuclear matrix of adenocarcinoma
cells (Wang et al., 1995b).

Ultrastructurally, galectin-3 is found in nucleoplasm, interchromatin spaces

excluding interchromatin granules, and on the border of condensed chromatins (Hubert et

l3

al.,l995). The latter is considered as the site of mRNA synthesis and early events of pre-
mRNA splicing (Spector, 1996). Therefore, this result agrees with the notion that
galectin-3 is a pre-mRNA splicing factor. Most surprisingly, galectin-3 is also associated
with dense fibrillar components and periphery of nucleoli, in contrast to the exclusion of
galectin-3 fi'om the nucleoli under fluorescence microscope (Hubert et al., 1995). What
causes this discrepancy is still not clear.

The import of galectin-3 into the nucleus has been studied and preliminary data
showed that galectin-3 probably enters the nucleus through an active mechanism.
Transfection experiments revealed that FLAG-tagged constructs expressing the duplicate
repeat and triplicate repeat of the galectin-3 coding sequence can enter the nucleus as well
as FLAG-tagged galectin-3 (Tsay and Wang, unpublished results). Inasmuch as both
duplicate and triplicate repeats have molecular sizes exceeding the diffusion limit of
nuclear pores, about 40~60 kDa for globular proteins (Peters, 1986), we concluded that
some active mechanisms are required for the entry for these constructs. Whether there
exists an import signal for galectin-3 itself still awaits further investigation.

Galectin-l is also present in nuclear structures that contain splicing factors.
Confocal microscopic experiments showed that galectin-l is also localized in nuclear
structures with Sm proteins as well as SC35, which supports the notion that it participates
in pre-mRNA splicing (V yakarnam et al., 1997). It will be interesting to know whether
its association with these structure is also dependent on RNAs. If the association is also
sensitive to RNase treatment, this implicates that the CRD is probably responsible for the

binding activity for both galectin-1 and galectin-3. However, if no RNA dependence is

l4

seen for galectin-1, it will support the notion that the N-domain of galectin-3 perhaps

plays an important role in association with nuclear matrix.

BIOLOGICAL FUNCTIONS

Galectins were initially found in the quest for proteins that recognize cell surface
carbohydrates, based on the presumption that they perhaps participated in cell-cell
interactions (Barondes, 1997). However, there is accumulating evidence indicating that
they have many additional functions, both intracellularly and extracellularly. Whereas
most activities are related to their saccharide binding activity, some are not necessarily
dependent on this characteristic property.

Galectins are involved in programmed cell death. Galectin-l has been found
to induce apoptosis in activated T cell, (Perillo et al., 1995) and thymocytes (Perillo et
al., 1997). Apoptosis is an important mechanism for the production of
irnmunocompetent cells and for the termination of an immune response. In activated T
cells, galectin-1 induces apoptosis via a Pas-independent pathway, and CD45 appears to
be the cell surface receptor that interacts with galectin-1. N-glycans on CD45 may be
essential for galectin-l binding, since the induction by galectin-l can be inhibited by
swainsonin, an inhibitor for N-linked oligosaccharide processing (Perillo et al., 1995).
Because most of galectin-1 is in dimeric form at the concentration (10 11M) that induces
apoptosis, dimeric formation was proposed as a requirement to elicit the reaction (Perillo

et al., 1995). Moreover, susceptibility to galectin-l-induced apoptosis is cell cycle-

15

dependent. Thymocytes at G1 phase are less sensitive to apoptotic change induced by
galectin-1 than at other phases (Perillo et al., 1997).

A tandem-repeat type galectin, galectin-9, also was found to induce cell death in
thymocytes (Wada et al., 1997). The induction is cell type-specific, since hepatocytes are
resistant to the stimulation. Lactose can suppress the response, suggesting that sugar-
lectin interaction is involved in this process.

Galectin-3 did not elicite simliar response on thymocytes (Wada et al., 1997). In
contrast, it protects leukemic T cells from the apoptotic response induced by anti-Fas
antibody and staurosporine (Yang et al., 1996). Interestingly, galectin-3 contains NWGR
motif seen in the Bel-2 family, which is involved in heterodimer formation between Bel-2
(a suppresser of cell death) and Bax (a cytotoxic protein) (Yin et a1, 1994; Hanada et al.,
1995). Irnmunoprecipitation analyses showed that galectin-3 indeed complexes with Bcl-
2 (Yang et al., 1996), but its significance in the apoptotic pathway remains to be
determined

Galectin-l participate in axon growth and guidance for olfactory neurons.
Galectin-l is expressed by primary Olfactory neurons, olfactory nerve glial cell, and
second-order neurons in the olfactory bulb (Puche and Key, 1995). In vitro, recombinant
galectin-1 and its ligand can specifically stimulate neurite growth by cultured olfactory
neurons, which were determined by increased neurite numbers as well as increased
neurite lengths. (Puche et al., 1996). When intemeuron connection was examined in
- transgenic null mice lacking galectin-1 (Poirier and Robertson, 1993), connections
between neurons in the nasal cavity and the olfactory bulb were defective. A

subpopulation of axons failed to navigate to their target site in the olfactory bulb (Puche

16

et al., 1996). This is the first phenotype ascribed to the lack of galectin-1 and indicates
that galectin-1 has a role in the guidance of axon growth of olfactory neurons.

Galectin-l and galectin-3 are pre-mRNA splicing factors. The observation
that saccharide ligands for galectins inhibited splicing in a cell-free assay first implicated
that galectins may participate in pre-mRNA splicing (Wang et al., 1991; Dagher et al.,
1995). When galectins were selectively depleted from nuclear extracts of HeLa cells
using lactose-agarose affinity adsorption, the splicing activity was lost, showing that
galectins are essential for pre-mRNA splicing. For these galectin-flee extracts, either
recombinant galectin-3 or galectin-l was able to restore its splicing capacity, and this
restored activity is still subjected to saccharide inhibition (Dagher et al., 1995;
Vyakamam et al. , 1997). This result indicated that either galectin-3 or galectin-1 is
sufficient to complement the splicing activity and also showing the importance of CRD.
The latter notion is further supported by the observation that C-domain of galectin-3
alone also reconstitutes the splicing activity (V yakarnam et al., 1997). When these two
galectins were specifically removed from nuclear extracts by dual immunodepletion, the
results were the same as those using lactose affmity adsorption. This argues that galectin-
] and galectin-3 are the two major galectins present in the nuclear extracts.

How do galectins participate in the splicing pathway? Assembly of splicing
substrates into splicing complexes in cell-free assay (figure 3) can be monitored by a
ribonucleoprotein (RNP) gel electrophoresis system (Zilhnann et al., 1988). When
splicing substrates were treated with galectin-free nuclear extract, only H complexes were
Observed on the gel. When recombinant galectin-3 was added to restore the splicing

activity, it also enhanced the formation of advanced spliceosomes including A complex

Figure 3. Schematic representation of the involvement of small nuclear
ribonucleoprotein particles (snRNP) and non-snRNP proteins in pre-mRNA splicing
(modified from Moore et al., 1993). Pre-mRNA (top line), containing two exons
separated by an intron, enters splicing complexes with snRNPs and exits as ligated
mRNA (bottom line). Nascent pre-mRNA is bonnd by a multitude of hnRNP proteins
immediately after transcription in vivo (Dreyfuss et al., 1993; Weighardt et al., 1996),
forming H complex. E complex, the commitment complex in mammalian extracts,
contains both U1 and U2 snRNPs, although U2 snRNP has a weaker affinity to the
substrate at this stage (Michaud and Reed, 1991). Transition from E complex to A
complex is rapid and requires ATP hydrolysis (Liao et al., 1992). The SR protein family
of splicing factors are involved in this conversion (V alcarcel and Green, 1996). Then,
U4/U6.U5 tri-snRNP complex enter spliceosome to form complex Bl. After structural
arrangement, complex B1 is transformed into complex B2. Complex C1 carries out the
cleavage at the 5' splice site and formation of the lariat intermediate, whereas complex C2

in the spliceosome immediately following exon ligation.

18

——A3’-n PRE-mRNA

+ HnRNPs
m—Al— HcomeLex
U51 * *r 1121 ”.2"

1112“;

m” —Ar Pyt— E COMPLEX

SFSI:SF3b SC35
ATP L33 ca

A COMPLEX

B1 COMPLEX

32 COMPLEX

C1 COMPLEX

C2 COMPLEX

 

19
as well as B1/B2 and C1/C2 complexes (Dagher et al., 1995; Zillmann et al., 1988). This

result implied that galectin-3 is at least required for the progression fiom H complex to A
complex. It is not known whether depletion of galectins blocks the conversion from H to
B complex, or from H to A complex (Figure 4).

Considering that CRD alone is sufficient to restore the splicing activity of
galectin-depleted nuclear extracts, it appears that carbohydrate-recognition domain
already contains sufficient elements mediating the splicing reaction. A legitimate
, question is whether carbohydrate-binding activity is required for splicing activity of
galectins. Recent data show that the galectin-1 mutant without sugar-binding activity is
still able to restore the splicing activity, although the saccharide inhibition is no longer
observed (Dagher and Patterson, unpublished results). This indicates that sugar binding
activity is, in fact, dispensable for splicing activity, arguing against the model that a
sugar-lectin interaction is required for progression of spliceosome. A plausible
explanation is that inhibition of splicing by saccharide perhaps reflects that the impact on
the CRD conformation by saccharide binding. It has been shown that galectin-3 CRD
have conformations of different thermal stabilities in the presence or absence of ligands
(Agrwal et al., 1993). It is possible that these two conformations represent active and
inactive forms for splicing activities of CRDs and lactose ligands act as a switch for their
interchange. Without ligand, the active form participates in splicing reaction; with bound
ligand, the active form becomes converted into the inactive form and thus CRD loses its
splicing activity (Figure 4). Whether there exits a switch in intact cells will be another

interesting issue.

20

Figure 4. Roles of galectin CRD in pre-mRN A splicing. Galectin CRD participates in
the conversion from H complex to A complex (Dagher et al., 1995). Depletion of
galectins by lactose-affmity adsorption or immunopresipitation inhibits the progression of
spliceosome formation. On the other hand, saccharide ligand induces a conformational
change on CRD structures, which results in inactivation of galectins and thus inhibition

of splicing reaction.

21

m—A-PY—m

l HnRNPs

H COMPLEX

IMMUNODEPLETION
LACTOSE ADSORPTION

 

® ACTIVE

 

+ LACTOSE
- LACTOSE

--. flflflflflflfl

     

SFSaISF3b

A COMPLEX

. INACTIVE

 

22
INTRODUCTION TO NUCLEAR TRANSPORT

Nucleocytoplasmic transport plays a fundamental role in coordinating the
functions of the nucleus and cytoplasm (Nigg, 1997; GOrlich and Mattaj, 1996). It is
mediated by a large molecular structure that spans the nuclear envelope called the nuclear
pore complex (Davis, 1995). Proteins that function in the nucleus are synthesized on free
ribosomes in the cytoplasm and imported into the nucleus. RNAS that function in
cytoplasm are transcribed and processed in the nucleus and exported into the cytoplasm
(GOrlich and Mattaj, 1996). Both processes occur through nuclear pore complexes. In
the following section, I will discuss the structure of nuclear pore complex, the pathways

for nuclear import, as well as the general properties of nuclear export.

NUCLEAR PORE COMPLEX

Nuclear pore complexes (NPC) have a mass of 125 megadaltons and contain
about 100 different polypeptides. Characteristic features of many vertebrate
nucleoporins (nuclear pore proteins) are modifications with O-linked N-
acetylglucosamine (GlcNAc) and the presence of short degenerate repeats, such as FXFG
repeats and GLFG repeats. The NPC contains a passive diffusion channel about 9 nm in
diameter (Peters, 1986). Small proteins with sizes less than 40-60 KDa can diffuse freely
through the pore. Proteins above the diffusion limit can enter the nucleus only via an
active mechanism. Nevertheless, small nuclear proteins generally enter the nucleus via

active mechanism rather than by diffusion.

23

Central
Plug

 
 
 
 
 
  
   
 
  

Cytoplasmic
Filament

Cytoplasmic

Ring

' . Lumenal
~ - Domain

Nuclear
Basket

Terminal

Ring
Figure 5. Schematic respresentation of a consensus model of the membrane-bound
nuclear pore complex. Nuclear pore complex contains four major elements: spoke-ring

assembly, central plug, cytoplasmic filaments, and nuclear basket (Bastos et al., 1995).

24

F our basic elements of structure have been observed for nuclear pore complex
(NPC) (Akey and Radermacher, 1993) (Figure 5). The main body is a spoke-ring
assembly anchored on a specialized region of nuclear envelope. This structure consists of
nuclear and cytoplasmic rings connected to each other by two sets of eight spokes . A
central plug sits within the aqueous channel formed by the spoke-ring assembly.
Peripheral components includes eight short cytoplasmic filaments that extend into the
cytoplasm and nuclear basket that project into the nucleus.

Many vertebrate nuclear pore proteins are O-linked glycoproteins that
contains characteristic peptide repeat motifs. “A family of 10-20 nucleoporins has
been identified as proteins that bind to wheat germ agglutinin (WGA) (Davis and Blobel,
1987). WGA binding is due to the extensive modification with O-linked N-
acetylglucosarnine moieties on the NPC structure (Hart et al., 1988). cDNAs encoding
six members (p62, POM121, NUP153, NUP214/CAN, NUP98 and NUP358/RanBP2) of
O-linked glycoprotein family of nucleoporins have been isolated in vertebrates (Table 2).
The deduced amino acid sequences showed that each member contains multiple copies of
XFXFG repeats, where X is any amino acid with a small or polar side chain. This motif
is also seen in a few yeast nucleoporins, like Nuplp (Bogerd et al. , 1994; Belanger et al. ,
1994) and Nup2p (Belanger et al., 1994). A GLFG motif, which is fi'equently seen on
yeast nucleoporins, has only been found on NUP98 in vertebrates.

Immunoelectron microscopy has been used to determine locations of these
glycoproteins within NPC. Namely, p62 is found both at cytoplasmic and nuclear sides
(Panté et al., 1994); POM21 is on membrane domain of NPC (Hallberg et al., 1993);

NUP358 as well as NUP214 are mapped to cytoplasmic filaments (Boer et al., 1997; Wu

25

Table 2. Vertebrate nuclear pore proteins

 

Name

Location

O-linked glycoprotein family

p62

POM121

NUP153

NUP2 14/
CAN

NUP98

NUP358/
RanBP2

Others
gp2 10

NUP107
NUP155

TPR/
p270
NUP180

NUP84
(NUP88)

hCRMl
(1 12kDa)

Central region; both cytoplasmic
and nuclear (Panté et al., 1994)

Pore membrane domain

Terminal ring of the nuclear basket
(Panté et al., 1994)

Cytoplasmic filaments (Panté et
al., 1994) and nucleoplasmic face
(Boer et al., 1997)
Nucleoplasmic side

Cytoplasmic fibers

Pore membrane domain

?
Both nuclear and cytoplasmic

Cytoplasmic (7) and nuclear
filaments (Cordes et al., 1997)
Cytoplasmic ring and filaments

Cytoplasmic face

Properties and possible functions

XF XF G repeats at N-tenninus and heptad
repeats/coiled-coil structure at C-domain.; complexed
with p58 and p54; essential for NPC function; multiple
O-GlcNAc and sialic acids (Snow et al., 1987; Starr et
al., 1990; Emig et al., 1995; Buss et al., 1995))

XFXF G repeats at C-terminus; integral membrane
protein; membrane anchor; multiple O-GlcNAc
(Hallberg et al, 1993)

XFXFG repeats at C-tenninus; four Zn fingers;
homodimerization; multiple O-GlcNAc ; involved in
mRNA export (Sukegawa and Blobel, 1993; Bastos et
al., 1996))

XFXF G repeats at C-terminus; leucine zipper motif;
complexed with p75; putative oncogene; docking site
for karyopherins (Kraemer et al., 1994)

GLFG, FG, FXFG repeats at N-terminus; docking sites
for karyopherins (N-terminus); related to DNA
replication; multiple O-GlcNAc; involved in RNA
export but not protein import (Powers et al., 1995;
Radu et al., 1995; Powers et al., 1997)

XFXFG, F G motifs; four Ran binding sites; leucine-
rich domain; cyclophilin A homologous domain;
docking sites for karyopherins and Ran (Y okoyarna et
al., 1995; Wu et al., 1995)

Type I integral membrane protein; membrane anchor;
high mannose N-linked oligosaccharides on lumenal
domain (Gerace et al., 1982; Wozniak et al., 1989;
Greber et al., 1990; Wozniak and Blobel, 1992)
Leucine zipper at C-terminus (Radu et al., 1994)
Multiple potential phosphorylation sites (Radu et al.,
1993)

Coiled-coil motif at central domain; activation of
oncogenic kinases (Byrd et al., 1994)

Recognized by serum form patients with autoimmune
diseases (Wilken er al., 1993)

Coiled-coil domain at C-terminus; bound with
Nup214; non-glycosylated; no repeat motif; CAN-
dependent NPC localization (Bastos et al., 1997;
Fomerod et al., 1 177))

Localized at pores and nucleoplasm; karyopherin 13-
like domain (Fomerod et al., 1997)

26
et al., 1995); and NUP98 and NUP153 are on the nuclear side of NPC (Panté et al., 1994;

Radu et al., 1995). In other words, nucleoporins with XFXF G repeats can be along the
cytoplasmic-nuclear axis of NPC.

The O-linked glycoproteins are essential docking sites for nuclear transport.
The majority of these proteins bind to the karyopherins, a family of transport factors,
directly, and have be proposed to form an array of sites for mediated docking of transport
substrates (figure 6) (Radu et al., 1995; Wu et al., 1995). For example, NUP358 has been
considered as the site where protein import complex is initially assembled (Wu et al.,
1995), containing docking sites for Ran GTPase as well as karyopherin Oil/01243
complexes. In fact, more and more data suggest that anchoring sites on these
nucleoporins are shared by all of the nuclear import pathways and export pathways. This
is exemplified by the fact that the B—type karyopherins are all capable of binding to these
glycoproteins, and their binding is competitively inhibited by each other (Bonifaci et al. ,
1997; Yaseen and Blobel, 1997). Moreover, it was recently shown that all of the signal-
mediated pathways were inhibited, when these glycoproteins in NPC were masked by
mutants of karyopherin Bl (Kutay et al., 1997).

Other nucleoporins. Sequence information is also available for several other
vertebrate nucleoporins (Table 2). None of them contains XFXFG or GLF G repeats nor
have O-linked GlcNAc. Some contain the sequence motifs that are likely to mediate
protein-protein or protein-nucleic acid interactions, such as zinc fingers, leucine zippers

and coiled-coil motifs.

27

    
 

NUP358*
NUP214*
NUP84
NUP180
gp210
*
NUP155 P62
Pomzl‘ NUP214*
NUP98*
Tpr/p27 *
NUP153

Figure 6. Diagram summarizing localizations of characterized nucleoporins within
the nuclear pore complex. Nucleoporins that have been shown to bind to karyopherin
complexes are placed at the right side. These proteins may serve as docking sites as import
complexes move along the nuclear pore. Asterisk indicates nucleoporins that have XFXFG

repeats as well as O-linked GlcNAc modifications.

28

Most yeast nucleoporins also have characteristic repeat motifs seen in
vertebrate homologues but are not modified by sugar residues. Some yeast
nucleoporins contains XFXFG repeats that are the signature motif of NPC O-linked
glycoproteins of vertebrates. These yeast proteins, however, appear not to be modified by
GlcNAc (Kalinich and Douglas, 1989), which is consistent with the fact that WGA does
not block nuclear protein import in in vitro import assay (Kalinich and Douglas, 1989).
A second family of nucleoporins is characterized by the repeat GLFG. It appears that
yeast and vertebrate NPCs have some differences, considering that the basic structure

units and posttranslational modifications are not entirely identical.

NUCLEAR IMPORT

As mentioned, molecules with sizes above the exclusion limits (40-60 kDa)
primarily enter the nucleus in an active way. Usually, this kind of transport is selective
and signal-dependent. Proteins or RNAs bearing import signals are recognized by
specialized receptors and then targeted to nuclear pores, where these molecules are
translocated across the nuclear envelope. Accumulating evidence implies that specificity
of a multitude of pathway arises at the earlier steps, i.e., signal-receptor recognition.
Different import pathways appear to use the same set of nucleoporins to dock import
substrates and translocate them across the nuclear pores (Bonifaci et al., 1997; Yaseen
and Blobel, 1997). Therefore, I would like to discuss the import pathways based on the

pairing between import signals and their receptors (Table 3).

29

Table 3. Signals and Receptors Involved in Nuclear Import

 

 

Receptor Import Signal Example Sequence Substrates
Karyopherin (11 Nuclear Localization Nuclear proteins
(Importin-a/NPI-l Signal PKKKRKV SV40 T antigen
SRPl/NTPCSS/ Classical signal KRPAAIKKAGQAKKKK Nucleoplasmin
Kap60p) Bipattite signal
Karyopherin (12 Nuclear Localization Nuclear proteins
(Rchl) Signal PKKKRKV SV40 T antigen
Classical signal KRPAAIKKAGQAKKKK Nucleoplasmin
Bipattite signal
Karyopherin a3 ? ? ?
Karyopherin B IBB domain of RMRKFKNKGKDTAELRR Karyopherin or
(Kap97p) karyopherin a RRVEVSVELRKAKKDE
QILKRRNV
Transportin M9 sequence NQSSNFGPMKGGNFGGR hnRNP A1
(Karyopherin B2 SSGPYGGGGQYFAKPR
/Kap104p) NQGGY
Karyopherin B3 ? ? Ribosomal proteins?
? KNS sequence YDRRGRPGDRYDGMVGF hnRNP K
SADETWDSAIDTWSPSE
WQMAY
? Pro, Gly-rich domain N-tenninus (aa 1-206) ANNEXIN XI
Snurportin Trimethylguanosine cap U snRNAs

 

30

Karyopherin a/Bl (011 B1 and a2Bl) mediated pathway for common nuclear
proteins is the best characterized nuclear import pathway. Most knowledge of this
pathway originates from studies in Xenopus and mammalian systems. Experiments based
on microinjection into Xenopus oocytes discovered the two steps of this import process:
energy-independent docking to the NPC and energy-dependent translocation through the
pore (Richardson et al., 1988; Newmeyer and Forbes, 1988). The major breakthrough
was the development of an in vitro system based on cultured cells treated with digitonin
to selectively permeabilize the plasma membrane (Adam et al., 1990). Digitonin-
permeabilized cells retain the capacity to import nuclear proteins with the supplement of
fractionated cytosol. From these cytosolic fractions, required factors for nuclear import
were purified.

Nuclear import of NLS-containing proteins has been distinguished as two
steps: docking and translocation (Figure 7). First, the import substrates binds via its
NLS to the karyopherin or, either 011 or 012, in the cytoplasm (GOrlich et al., 1994;
Imamoto et al., 1995), which also augments the interaction between karyopherin or and
B1 subunits (Moroianu et al., 1996). Karyopherin B1 contains binding sites with repeat-
containing nucleoporins, and mediates the docking of the complex onto the nuclear pores
in an energy-independent manner (Morioanu et al., 1995). The import complex is
subsequently translocated through the pore by an energy-dependent mechanism
(Richardson et al., 1988; Newmeyer and Forbes, 1988), which is probably achieved by
sequential docking-undocking process mediated by Ran, a Ras-like GTPase (Melchior et

al., 1993; Moore and Blobel, 1993; Kutay et al., 1997), and NTF2 (Moore and Blobel,

 

Figure 7. Nuclear import of NLS-containing proteins has two distinct steps: docking

and translocation. Docking requires only karyopherina and B, and does not need GTP.
Translocation is mediated by Ran GTPase and its cofactor NTF2, which requires GTP

hydrolysis.

32
1994; Paschal and Gerace, 1995). After being transported to the nuclear side of NPC, the

import complex is dissociated. As the karyopherin B1 was promptly exported by an
NBS-mediated pathway (Iovine and Wente, 1997), the karyopherin (Jr-substrate complex
reaches the nucleoplasm (Moroianu et al., 1995).

Ran GTPase cycle is coupled with the translocation of karyopherin complex.
Like other GTPases, Ran has a set of effectors that modulate its GTPase activity. The
only known guanine nucleotide exchange factor for Ran is a nuclear protein RCCl
(Bischoff and Ponstingl, 1991); the only known Ran GTPase activating protein is
cytoplasmic (Hopper et al., 1990; Bischoff et al., 1995). Based on distinct locations of
these two effectors, it appears that GTP exchange occurs only in the nucleus and GTP
hydrolysis is only in the cytoplasm.

The Ran GTPase cycle. Koepp and Silver (1996) proposed that Ran moves out
of the nucleus in a GTP-bound form and it enters the nucleus at the GDP-bound state
(Koepp and Silver, 1996) (Figure 8A), assuming that Ran-GTP is enriched by nuclear
RCCl in the nucleus and Ran-GDP is formed in cytoplasm due to augmented GTPase
activity by RanGAP. This model also predicted that Ran-GDP moves toward cytoplasm
and Ran-GTP heads for the nuclear side so that the cycle can be completed (Figure 8).
Some of the experimental data are consistent this model:

1. Ran-GDP has been found to be required for nuclear import (Weis et al., 1996).
2. When RanGAP was microinjected into the nuclei, nuclear export of NES-

containing protein was blocked (Richards et al., 1997).

33

Figure 8: (A) A model illustrating Ran GTPase cycle. Two effectors for Ran are at
different compartments. RCCl is localized in the nucleus and RanGAP is in the
cytoplasm. In order the complete the cycle, Ran-GDP flows toward the nucleus and Ran-
GTP move toward cytoplasm. (B) Coordination of protein import by Ran GTPase
cycle. Due to the compartrnentation of RCCl and RanGAP, Ran-GDP and Ran-GTP are
enriched at different sides. Ran-GDP is more abundant at cytoplasmic side, where it
promotes the docking complex for protein import. Ran-GTP is more abundant at the
nuclear side, where it stimulates the dissociation of karyopherin-substrate complex. The

net effect is that nuclear protein is transported from cytoplasmic side to the nuclear side.

 

 

35
3. GTP-bound Ran mutant lacking GTPase activity is still able to mediate nuclear

export, indicating that Ran-GTP but not GTP hydrolysis directly participate in

nuclear export (Richards et al., 1997).

4. When nuclear import was blocked by nonhydrolyzable GTP analogues, Ran was

specifically localized at the cytoplasmic surface (Melchior et al., 1995).

Current data indicates that Ran GTPase cycle coordinates the import pathway. It
has been shown that karyopherin 01/ B1 and imported substrate form a stable complex in
the presence of GDP-bound Ran, but karyopherin a-ligand is released from the complex
by Ran-GTP (Rexach and Blobel, 1995; Kutay et al., 1997). Since the Ran-GTP is
mainly nuclear and Ran-GDP is enriched in cytoplasm according to the model, the Ran-
GDP mediates the import complex formation at the cytoplasmic side and Ran-GTP will
dissociate the substrate from karyopherins in the nucleus (Figure 8B).

In summary, the distinct distribution of RCCl and RanGAP across the nuclear
envelope enriches GDP- and GTP-bound forms of Ran at cytoplasmic side and nuclear
side respectively. This, in tum, potentially helps to establish the unidirectional
movement of nuclear proteins (Figure 8B).

NTF2 assists the docking of Ran-GDP as well as import complexes to the
repeat-containing nucleoporins. NTF2 was first identified as a required factor for
nuclear import, complementing the functions of Ran (Moore and Blobel, 1994). In yeast,
NTF2 has specific affinity to nucleoporins containing FXFG repeats, like Nuplp, Nsplp,
Nupl45p , Nup57p and Nup36p. Stronger binding with some of these nucleoporins, like

Nup36p requires the presence of Ran-GDP, Kap95B and Kap60a. Therefore, NTF2 has

36

been proposed as the regulator that coordinates Ran-dependent association and
dissociation reactions underlying nuclear import (Nehrbass and Blobel, 1996). A similar
interaction between karyopherins, Ran, nucleoporin, and NTF2 has also been observed in
vertebrate cells (Paschal and Gerace, 1995).

Transportin-M9 sequence mediated pathway is responsible for the nuclear
import of hnRNP Al and its related proteins. HnRNPs are a set of proteins that bind
to hnRNAs. They are involved in many aspects of RNA metabolism, including RNA
processing and transport (Dreyfuss et al., 1993). In hnRNP A1, there exists a stretch of
amino acids, designated as M9 (Siomi and Dreyfuss, 1995), that is responsible for nuclear
shuttling property of hnRNP A1 (Piflol-Roma and Dreyfuss, 1993). Recently, its nuclear
import receptor has been identified in a variety of organisms. In vertebrates, it was
designated as transportin (Pollard and Dreyfuss, 1996), MIP (M9 Interacting Protein)
(Fridell et al., 1997), and karyopherin B2; in yeast, it is named as Kap104p (Aitchison et
al., 1996). This receptor is homologous to karyopherin B1 and specifically mediates
nuclear import of hnRNP Al, but not NLS-containing proteins (Pollard et al., 1996;
Fridell et al., 1997). Transportin, like karyopherin B1, binds to a set of nucleoporins
containing the characteristic peptide repeat motifs. It mediates the docking of hnRNP A1
onto the pore, but translocation also requires Ran GTPase (Bonifaci and Blobel, 1997).

There are a multitude of nuclear import pathways yet to be defined. For
example, KNS sequence in hnRNP K is responsible for its nuclear shuttling property,
and, interestingly, this sequence can dock proteins to the nuclear pore without the

supplement of cytosolic factors (Michael et al., 1997). What nucleoporin mediates its

37

docking still remains to be studied. Some other potential receptor proteins are also
identified, like karyopherin 1313 (Takeda et al., 1997) and karyopherin B3 (Yaseen and
Blobel, 1997), but no substrates have been characterized yet. There is also a report on the
identification of the receptor that mediates nuclear import of U snRNAs, which has a
trirnethylguanosine cap that interacts with a receptor, snurportin (Lilhrmann, 1997). This
receptor mediates nuclear import of U snRNPs in oocyte cells but only plays a minor role

in somatic cells (Marshallsay and Ltihrmann, 1994).

NUCLEAR EXPORT

The list of signals and receptors involved in nuclear export is relatively small,
compared to that for nuclear import (Table 4). Discovery of the cap-binding complex
(Izaurralde et al., 1995) provided the first chance to examine the molecular mechanisms
underlying RNA export and to explore the interplays between nuclear import and export.
Nuclear export signals have also been recently identified in several nuclear shuttling
proteins (Gerace, 1995).

Cap-binding complex (CBC) is involved in RNA export. Monomethylated cap
structures have been found to participate in several steps of RNA metabolism, including
nuclear export (Harnm and Mattaj, 1990), initiation of translation (Sachs et al., 1997) and
RNA processing (Izaurralde et al., 1994). Microinjection of cap analog, the dinucleotide
m7GpppG, was shown to specifically inhibit nuclear export of U1 snRNA. For mRNAs,
a monomethylated cap conferred a higher export rate than a dimethylated or trimethylated

cap structure (Harnm and Mattaj, 1990; Jarmolowski et al., 1994), although the

38

Table 4. Signals and Receptor Involved in Nuclear Export

 

 

Receptor Export signal Example Sequence Substrate

Cap-binding 7-methylguanosine cap mGpppG, mGpppA U snRNAs; mRNAs

complex

hRIP Nuclear Export Signal LPPLERLTL (Rev) Nuclear proteins

(NES) LALKLAGLDI (PKI)

? M9 sequence NQSSNFGPMKGGNFGGR hnRNP A1
SSGPYGGGGQYFAKPR
NQGGY

? KNS sequence YDRRGRPGDRYDGMVG hnRNP K
FSADETWDSAIDTWSP

SEWQMAY

 

39

monomethylated cap was not essential for their export (Izaunalde et al., 1997). A cap-
binding complex (CBC) was identified from nuclear extracts, which consists of two
subunits: CBP20 and CBP8O (Izaurralde et al., 1992; Izaurralde et al., 1995). Antibodies
raised against recombinant CBP20 indeed inhibited the nuclear export of U snRNAs,
demonstrating that this complex is an nuclear export factor for RNA (Izaurralde et al.,
1995).

Although CBC seems not to be required for export of mRNAs, it does accompany
mRNPs through nuclear pores on a particular pre-mRNA particle, the Blabiani ring (BR)
RNP. BR particle is the product of the Balbiani ring (BR) gene in the salivary glands of
the dipterin Chironomus tentans (Visa et al., 1996), whose giant size (35-40 kb) makes it
possible to be visualized under electron microscope. CBC binds to the BR transcript as it
was transcribed, and accompanies this RNP particle to travel through the nucleoplasm
until it meets nuclear pores. During translocation through the channel, CBC remains at
the 5’ end of the particle and is not detached until the RNA molecule extends into the
cytoplasm (Figure 9) (Visa et al., 1996). Thereafter, it appears to be released from the
BR RNP immediately, probably reentering the nucleus. Recent data showed that
karyopherin a-Bl pathway is responsible for the nuclear import of this NLS-containing
CBC (GOrlich et al., 1996).

Nuclear export signals (NES) in certain nuclear shuttling proteins are
responsible for their rapid exit from the cell nucleus. Nuclear export signals have
been identified in several nuclear shuttling proteins. They were initially found in the

HIV-1 Rev protein (Wen et al., 1995; Fischer et al., 1995) and in the Protein kinase A

40

BRRNP ____.. —> 40—. ._____+ 6,,

 

Figure 9: The nuclear export of Balbiani Ring RNP particle. Three pre-mRN A splicing
factors, hrp45, CBC20 and hrp36, travel with the transcript to different locations (modified

form Danholt (1 997)).

41
inhibitor (PKI) (Wen et al., 1995). Both are short sequences and enriched in hydrophobic

residues, especially leucines, with particular spacing. NBS-mediated export is energy and
temperature dependent and confers a rapid export property to heterologous proteins (Wen
et al., 1995). A receptor that interacts with this group of signals has been identified. The
NES receptor, hRIP (human Rev Interacting Protein) or RAB (Rev Activation domain
Binding protein), is a 59-kDa nucleoporin-like protein with FG peptide repeat motifs
(Fritz et al., 1995; Bogerd et al., 1995). It is localized in the nucleus as well as the
nuclear envelope (Fritz et al., 1995), and probably also cytoplasm (Bogerd et al., 1995).
Known nuclear proteins containing NES include ( 1) components for nuclear transport of
macromolecules, like Rev (Wen et al., 1995), Rex (Bogerd et al., 1996; Kim et al., 1996),
karyopherin B1 (Iovine and Wente, 1997), and RanBPl (Zolotukhin et al., 1997), and (2)
proteins involved in signal transduction pathways, like PKI (Wen et al., 1995), and
MAPK kinase (Fukuda et al., 1996).

M9-mediated pathway is related to nuclear export of mRNAs. Using a
heterokaryon cell system, it has been shown that hnRNP Al has a shuttling capacity
across nuclear envelope (Pifiol-Roma and Dreyfuss, 1993). The same conclusion is
derived from the observation on its yeast homologue Npl3p (Flach et al., 1994). A
domain was found to mediate the nuclear import and export activities, called M9
sequence (Siomi and Dreyfirss 1995; Michael et al., 1995). Although transportin is
considered as M9 receptor, it seems not to be responsible for its nuclear export (Pollard et
al., 1996; Bonifaci et al., 1997). When hnRNP Al was microinjected into either

cytoplasm or the nucleus, it indeed saturates the export pathway for mRN As, but not for

42
U snRNAs (Izaurralde et al., 1997) This inhibition is independent of M9 sequence and

may be exerted in cytoplasm, and the underlying mechanism is not clear at present. C
tentans hrp36, a homologue of hnRNP Al, is a component of BR RNP particle (Visa et
al., 1996). Its localizaiton has been studied using immunoelectron microscopy. This
study showed that it was added to BR RNA concomitant with transcription, and remained
on this RNP particle, while traveling through nucleoplasm. After leaving the nucleus,
hrp36 was still present in this RNP even when polysomes were assembled on BR
particles (Visa et al., 1996). Although hnRNP A] has extensive association with RNAS,
direct evidence indicating that it is essential for mRN A export is still lacking.

A novel signal conferring nuclear shuttling activities is found in hnRNP K.
This KNS sequence, like M9 sequence, is responsible for the nuclear export as well as
nuclear import property of hnRNP K1 , as demonstrated in a heterokaryon system
(Michael et al., 1997). The molecular mechanism for its export awaits further
investigation.

Some of the pre-mRNA splicing factors are exported with processed
transcripts. Both hnRNP A1 (Mayeda et al., 1992; Caceres et al., 1994) and CBC
(Izaurralde et al., 1994; Lewis et al., 1996) are well documented pre-mRNA splicing
factors. Although both of these proteins travel out of the nucleus with mRNAs, the
timing of their release form the RNP is quite different. CBC appears to be released from
the RNP complexes immediately following their arrival at cytoplasmic side of nuclear
pores (Visa et al., 1996a), while hnRNP A1 probably accompanies RNAs from
transcription to translation (Visa et al., 1996b). However, not every pre-mRNA splicing

factor is exported with mRNP complex. For example, the hrp45 protein, a SR protein of

43

C. tentans, is found to associate with nascent RNAs concomitant with their synthesis via
the binding with the exon. It is released from BR RNP particles in the nucleoplasm when
the particle make contacts with the nuclear pores (Alzhanova-Ericsson et al., 1996)
(Figure 9). Briefly speaking, different splicing factors seem to have distinct fates for
binding with the pre-mRNA/mRNA particles. Based on distinct timing of release, this
process seems to be a regulated process. How and why their binding with mRNPs is

terminated will be interesting subjects to explore.

CHAPTER II
NUCLEAR EXPORT OF GALECTIN-3 IN MOUSE 3T3 FIBROBLASTS:
I. PARAMETERS OF THE TRANSPORT IN

DIGITONIN-PERMEABILIZED CELLS

Yeou-Guang Tsay, Nancy Y. Lin, and John L.Wang

Department of Biochemistry, Michigan State University

East Lansing, MI 48824

44

45
SUMMARY

Galectin-3 is a galactose-Ilactose-binding protein (M, ~30,000), identified as a
required factor in the splicing of pre-mRN A. Immunofluorescence staining revealed that
galectin-3 distributes differentially between the nucleus and the cytoplasm, depending on
the proliferative state of the cells under analysis. Using digitonin-permeabilized mouse
3T3 fibroblasts, we provide evidence that galectin-3 is rapidly and selectively exported
from the nucleus. Although both phosphorylated and nonphosphorylated isoforms of
galectin-3 are found in the nuclear fraction, only phosphorylated galectin-3 is identified
in the exported fraction, implying that phosphorylation is important for the nuclear export
of the protein. In addition to galectin-3, we found that several other proteins are also
rapidly exported in the digitonin-permeabilized cell assay. These include galectin-l, Ran,
and karyopherin-B. In contrast, the Sm B and U1 A polypeptides of the small nuclear
ribonucleoprotein complexes (snRNPs) are retained in the nuclear residue during the
course of the same assay. The behavior of heterogeneous nuclear RNP (hnRNP) protein
A1 is somewhat intermediate, exhibiting very slow export. The rate of galectin-3 export
is decreased by cold temperature and by the addition of wheat germ agglutinin. More
strikingly, galectin-3 export can be inhibited, at least partially, by the simultaneous
addition of a peptide bearing a nuclear export signal plus a dinucleotide cap analogue
found at the 5’-end of mRNAs. These results suggest that galectin-3 may be exported in
association with RNP(s) containing monomethylated cap structure as well as

polypeptide(s) bearing nuclear export signals.

46
INTRODUCTION

Galectin-3 belongs to a family of widely distributed proteins that: (a) bind to B-
galactosides; and (b) contain characteristic amino acid sequences in the carbohydrate
recognition domain of the polypeptide (for reviews, see Barondes et al., 1994; Kasai and
Hirabayashi, 1996; Leffler, 1997). Of the presently known 10 members of this family,
galectins-1 and -3 have been shown to be factors involved in the splicing of pre-mRNA,
assayed in a cell-free system (Dagher et al. 1995; Vyakamam et al., 1997). This
conclusion is based on several key findings: (i) nuclear extracts (NE) derived from HeLa
cells, capable of carrying out splicing, contain galectins-1 and -3; (ii) saccharides that
bind the galectins with high affinity inhibit the cell-free splicing reaction; (iii) depletion
of both galectins from NE, either by lactose affinity adsorption or by double antibody
adsorption, resulted in the concomitant loss of splicing activity; (iv) depletion of either
galectin-1 or galectin-3, by specific antibody adsorption, failed to remove all of the
splicing activity and the residual activity was still sensitive to saccharide inhibition; and
(v) either galectin-1 or galectin-3 alone is sufficient to reconstitute, at least partially, the
splicing activity of NE depleted of both galectins. All of the results suggest that the
activities of galectin-1 and galectin-3 in the nucleus may be redundant (Patterson et al. ,

19971

A number of lines of evidence have been accumulated to indicate that galectin-3

is found in both the cytoplasm and nucleus of cells. Immunofluorescence staining, using

47

monoclonal, as well as polyclonal, antibodies specifically directed against galectin-3, was
carried out on formaldehyde fixed, Triton X-100 permeabilized 3T3 fibroblasts. There
was prominent labeling of the nucleus and variable staining of the cytoplasm (Moutsatsos
et al. 1986; Wang et al., 1995). This dual localization of galectin-3, in the nucleus and
cytoplasm, has been confirmed by immunoelectron microscopy (Hubert et al., 1995).
Moreover, the nuclear staining of galectin-3, at the light and electron microscopic levels,
was sensitive to treatment of the cells with ribonuclease A, but not to parallel treatment
with deoxyribonuclease I, implying galectin-3 is associated with RNA-containing nuclear
structures. Finally, the predominance of nuclear versus cytoplasmic distribution of
galectin-3 depended on the proliferative state of the cells: sparse, proliferating cultures
showed nuclear localization of galectin-3 whereas confluent, quiescent cultures yielded
mostly cytoplasmic staining (Moutsatsos et al., 1987). Galectin-3 expression and its
intracellular distribution have also been shown to vary along the crypt-tO-surface axis of
human colonic epithelia (Lotz et al., 1993). The protein is concentrated in the nuclei of
differentiated colonic epithelial cells. The progression from normal mucosa to adenoma
to carcinoma is characterized by a paradoxically distinct absence of galectin-3 in the

nuclei of adenoma and carcinoma cells.

These observations on the dual localization of galectin-3 raise the possibility that
the protein might shuttle between the cytoplasm and nuclear compartments. The rapidly
expanding list of proteins that exhibit this nucleocytoplasmic shuttling property includes:
(i) nucleolar proteins nucleolin and B23/No38 (Borer et al. , 1989); (ii) heat shock protein

hsp70 (Mandell and Feldherr, 1990); (iii) transcription factors such as the steroid

48
receptors (Guiochon-Mantel et al., 1991; Chandran and DeFranco, 1992) and TFIIIA

(F ridell et al., 1996); (iv) proteins of the heterogenenous nuclear ribonucleoprotein
complex (hnRNP) such as hnRNP A1, A2, E (Pifiol-Roma, and Dreyfuss, 1993) and
hnRNP K (Michael et al., 1997); (v) proteins involved in nucleocytoplasmic transport,
including karyopherin-/importin-B (Iovine and Wente, 1997), Ran-binding Protein
' (Richards et al., 1996), the HIV Rev protein (Wen et al., 1995; Fischer et al., 1995), and
Np13p (F lach et al., 1994; Lee et al., 1996); and (vi) various signal transduction proteins
and their regulators, such as the heat stable inhibitor of CAMP-dependent protein kinase
(PKI) (Wen et al., 1995) and mitogen-activated protein kinase kinase (F ukuda et al.,

1996)

Studies on nuclear protein import have identified some of the key components
that function in recognition, docking, and translocation steps of the process.
Karyopherin-/importin-Or is one subunit of the receptor that recognizes the nuclear
localization signal (NLS) and, along with karyopherin-B, mediates the docking of import-
competent substrates to the nuclear pore complex (GOrlich et al., 1994; Radu et al.,
1995). A GTP-binding protein Ran/TC4 and the associated factor, plO/NTF 2, are
required for the translocation of the substrate proteins through the nuclear pore complex
(Moore and Blobel, 1993; Melchior et al., 1993). These advances in our understanding of
nuclear protein import have been possible mainly as a result of the development of an in
vitro assay, using digitonin-penneabilized cells (Adam et al., 1990). This assay takes

advantage of the fact that digitonin treatment of cultured cells penneabilizes the plasma

49

membrane to macromolecular substrates while retaining the structural integrity of the
nuclear envelope. Thus, the nuclei retain the ability to transport and accumulate proteins
bearing the appropriate NLS. The system is particularly amenable to fractionation and

biochemical analysis of the components involved in the transport process.

We have taken advantage of this digitonin-permeabilized cell system to test the
possibility that galectin-3 might be exported from the cell nucleus, a requirement for a
nuclear protein to shuttle between the nucleus and the cytoplasm. In the present
communication, we report that, indeed, galectin-3 is rapidly and selectively exported
from the nucleus of digitonin-permeabilized cells. In the accompanying paper, we report

the results of our initial characterization of the exported components.

50
EXPERIMENTAL PROCEDURES

Cell Culture and Reagents

NIH mouse 3T3 fibroblasts were Obtained from the American Type Culture
Collection (Rockville, MD). The cells were grown as monolayers in Dulbecco’s
Modified Eagle Medium containing 10% calf serum, 100 U/ml penicillin, and 100 ug/ml
streptomycin at 37 °C in a humidified atomsphere of 10% C02. Routinely, the cells were
used in the export assay at a density of ~ 1.5 x 10‘ cells/cmz, about 30% of the saturation
density. The nuclear export assay was also carried out using 3T3 cells metabolically
labeled with 32PO4 (New England Nuclear, Boston, MA; 50 uCi/ml culture medium, 5 ml/

culture dish).

A rat monoclonal antibody was developed against the Mac-2 antigen (Ho and
Springer, 1982), which has been shown to be galectin-3 (Cherayil et al., 1989). The
antibody was prepared from serum-free cell culture supernatant derived from the
hybridoma line as described (Wang et al., 1995); this antibody preparation is hereafter
designated as rat anti-galectin-3. The polyclonal rabbit antiserum against recombinant rat
galectin-1 (rabbit anti-galectin-l) (Cooper et al., 1991) was a gift from Drs. Sam
Barondes, Hakon Leffler, and Doug Cooper (University of California, San Francisco).

The mouse monoclonal antibody 9H10/4B10 (Pinol-Roma et al., 1988), directed against

51
hnRNP Al (M, ~34,000), was a generous gift of Dr. Gideon Dreyfuss (University of

Pennsylvania) and is designated as mouse anti-hnRNP A1.

The following mouse monoclonal antibodies were purchased from Transduction
Laboratories (Lexington, KY): (a) mouse anti-Ran was used to detect the GTP-binding-
protein Ran (M, ~ 25,000), involved in nucleocytoplasmic transport; and (b) mouse anti-
karyopherin-B was used to detect the B-subunit (M, ~97,000) of the NLS receptor.
Human autoimmune serum anti-Sm, used to detect Sm B (M, ~29,000) representing a
core polypeptide of small nuclear ribonucleoprotein complexes (snRNPs), was purchased
from The Binding Site (San Diego, CA). Human autoimmune serum used to detect the A
polypeptide (M, ~33,000) specific to U1 snRNP, anti-U1 RNP, was also purchased from
The Binding Site. Rabbit polyclonal antibody against phosphoserine was purchased from

Zymed Laboratories (South San Francisco, CA).

Wheat germ agglutinin (WGA) was purchased from Sigma (St. Louis, MO).
Nucleotides corresponding to the monomethylated cap structures at the 5’-end of RNA
polymerase II transcripts, m7GpppA and m7GpppG, were purchased from Boehringer
Mannheim (Indianapolis, IN). The peptide NELALKLAGLDINKTG corresponds to
residues 34-50 of PKI and has been shown to serve as the latter’s Nuclear Export Signal
(NES), with the underlined leucine and isoleucine residues being critical for activity
(Wen et al., 1995). This peptide was synthesized by the Macromolecular Structural

Facility at Michigan State University and is hereafter designated the NES peptide.

52

The Nuclear Exmrt Assay in Digitonin-Perrneabilized Cells

The following protocol is described for a typical assay (Fig. 1) derived from one
tissue culture dish (28 cm2 surface area). Cells, at ~30% of confluent density, were
washed with transport buffer (TB; 20 mM Hepes, pH 7.3, 110 mM KOAc, 2 mM EGTA,
and 2 mM Mg(OAc)2) The cells were permeabilized in TB containing digitonin (1.75
11M) for 5 min. at 4 °C. After removal of the digitonin solution containing the soluble
cytosolic components, 3 ml TB, supplemented with 1 mM dithiothreitol, 5 ug/ml
aprotinin, 5 rig/ml pepstatin, and 10 ug/ml leupeptin, were added onto the permeabilized
cells and incubated for various times of the export assay (Fig. 1). The supernatant was
then collected as the Transported Fraction (TF). To 1.5 ml of TF, 15 1.11 deoxycholate
(10%, w/w) and 150 [.11 trichloroacetic acid (100%) were added and the mixture was
incubated at 4 °C for 30 min. The suspension was then centrifuged (12,000 x g; 15 min.)
and the precipitate was washed with acetone. The precipitate was resuspended in 40 111
1% deoxycholate, 0.1 N NaOH; 10 111 of 5X SDS-PAGE sample buffer was then added

and the entire sample was subjected to electrophoresis.

In each experiment, the corresponding nuclear fraction (NF) was harvested from
the same culture dish by solubilization in 50 ul of SDS-PAGE sample buffer. The

components of the nuclear residue were analyzed by SDS-PAGE.

53

Figure 1: Schematic illustration of the digitonin-permeabilized cell system for the
assay of nuclear export. Monolayer cultures of mouse 3T3 cells were treated with
digitonin and then incubated with transport buffer. The presence of galectin-3 in the
nuclear fraction was monitored by immunofluorescence and by Western blotting. The

exported nuclear proteins were analyzed by Western blotting.

 

54

 

Soluble Digitonin
cytosolic

Components

  

‘~.

I
~-‘---l”

r- Transport Buffer

 

At various times

  

\ Exported
1 Nuclear
\ ,’ components

’
~_—’

Nuclear Fraction Transported Fraction

55

One- and Two—dimensional Gel Electrophoresis

Proteins were resolved on one-dimensional gels, using 12.5% SDS-PAGE as
described by Laemmli (1970). For two-dimensional gels, nonequilibrium pH gradient
electrophoresis (NEPHGE) in the first dimension and SDS-PAGE in the second dimen-
sion were used as described by O’Farrell et al. (1977) with the following minor modifi-
cations. The NEPHGE gels were made using only pH 3-10 arnpholines and the electro-
phoresis was can'ied out at 400 V for 5 h. Standard proteins for two-dimensional gels

(BioRad, Richmond, CA) were used to track the pH gradient.

Western Blotting and Quantitation

For irnmunoblotting, proteins resolved by SDS-PAGE were transferred
electrophoretically onto Immobilon-P membranes (Millipore, Bedford, MA) in buffer
containing 25 mM Tris, 193 mM glycine, and 10% methanol. The membranes were
blocked for several hours at room temperature with 10% non-fat dry milk in T-TBS (10
mM Tris, pH 7.5, 0.5 M NaCl, 0.05% Tween-20). After two brief washes with T-TBS,
the membranes were incubated with primary antibody (rat anti-galectin-B, rabbit anti-
galectin-1, mouse anti-hnRNP A1, mouse anti-Ran, mouse anti-karyopherin-B, human
anti-U1 RNP and human anti-Sm) at room temperature for 1 h. The membranes were
rinsed twice and then washed three times with T-TBS (one 15-min. wash and two 5-min

washes). They were then incubated with the appropriate alkaline phosphatase-conj ugated

56

secondary antibody (goat anti-rat immunoglobulin, goat anti-rabbit immunoglobulin, goat
anti-mouse immunoglobulin, or goat anti-human immunoglobulin, all at 1:2000 dilution)
for 30 min. at room temperature. After rinsing twice and washing in T-TBS (one lS-min
wash and two 5-min washes), the blots were developed using colorimetric substrates for

alkaline phosphatase, nitroblue tetrazolium and 5-bromo-4-chloro-3-indolyl phosphate.

For quantitation, the blots were scanned using a Hewlett-Packard ScanJet 3C and
the DeskScan 11 program (Hewlett Packard, Palo Alto, CA). The image was then
analyzed using Image Quant (Molecular Dynamics, Sunnyvale, CA) and the pixel values

Of the individual bands were integrated to yield a value for intensity.

Immunofluorescence

For immunofluorescence microscopy, 3T3 cells were seeded onto glass coverslips
placed in 6-well (8 cm2/well) cluster dishes. In the basic protocol, cells were washed
twice with phosphate-buffered saline (PBS, 140 mM NaCl, 2.68 mM KCl, 10 mM
Na,HPO,, 1.47 mM KH2P04, pH 7.4). and fixed for 20 min. at 4 °C with 4%
paraforrnaldehyde in PBS. Cells were washed twice in PBS and then permeabilized with
0.2% Triton X-100 in PBS for 5 min. at 4 °C. After permeabilization, the cells were
again washed twice with PBS and incubated with 0.2% gelatin in PBS for 1 h at room

temperature.

57

Several variations of this basic protocol were also performed: (i) the cells were
treated with digitonin as in the export assay (1.75 11M; 5 min.; 4 °C), followed
immediately by fixation with paraformaldehyde and permeabilization with Triton X-100
as described above; (ii) the cells were treated with digitonin as in the export assay, fixed
with paraformaldehyde but without the additional permeabilization with Triton X-100;
and (iii) the cells were permeabilized with digitonin, incubated with TB for 30 min.,
followed by fixation with paraformaldehyde and permeabilization with Triton X-100 as

described for the basic protocol.

In all cases, the cells, after the “blocking” incubation in 0.2% gelatin in PBS, were
washed in T-TBS. They were then incubated with rat anti-galectin-3 (250 ug/ml in T-
TBS containing 0.2% gelatin) for l h at 4 °C. The cells were washed three times (15 min.
each) with T-TBS and then incubated with fluorescein-labeled goat anti-rat
irnmunoglobulin (10 ug/ml in T-TBS containing 0.2% gelatin) for 30 min. at room
temperature. The coverslips were washed three times in T-TBS (15 min. each) and then
mounted in Perma Fluor (Lipshaw Immunon, Pittsburgh, PA) on glass slides.
Fluorescence staining was analyzed using a Meridian Instruments (Okemos, MI) Insight

confocal laser scanning microscope.

58
RESULTS

Exmrt of Galectin-3 from Nuclei of Digitonin-permeabilized cells

The digitonin-penneabilized cell system, used to assay the import of protein
bearing NLS (Adam et al., 1990; Moore and Blobel, 1993) was adapted to analyze the
export of galectin-3 from the nucleus. The presence of galectin-3 in the nuclear residue
was monitored by immunofluorescence and by Western blotting of NF. In parallel, TF,

containing the exported nuclear components, was also analyzed for galectin-3 (Fig. 1).

In accord with our previous observations (Laing and Wang, 1988; Wang et al.,
1995), mouse 3T3 fibroblasts, fixed with paraformaldehyde and permeabilized with
Triton X-100, yielded prominent nuclear and discernible cytoplasmic staining with anti-
galectin-3 (Fig. 2A). When 3T3 cells were first treated with digitonin to permeabilize the
plasma membrane, then fixed and treated with Triton X-100 to permeabilize the nuclear
membrane, a similar staining pattern of the cell nucleus was observed (Fig. 2B).
Cytoplasmic galectin-3 was lost upon permeabilization of the plasma membrane. The
galectin-3 detected by this procedure (Fig. 2B) is inside the nucleus. This conclusion is
based on the observation that digitonin-permeabilized cells, fixed with paraformaldehyde
but without Triton X-100 permeabilization of the nuclear membrane to allow antibody
accessibility, yielded no staining (Fig. 2C). When the digitonin-permeabilized cells were
incubated in TB for 30 minutes prior to fixation and Triton X-100 permeabilization,

galectin-3 staining was diminished drastically (compare Fig. 2D with Fig. 2B). These

59
.-
.-

Figure 2: Comparison of the immunofluorescence staining for galectin-3 in 3T3
cells before and after differential permeabilization. (A) Cells fixed with
paraformaldehyde (4%; 20 min; 4 °C) and permeabilized with Triton X-100 (0.2%; 5 min;
4 °C); (B) Cells permeabilized with digitonin (1.75 11M; 5 min; 4 °C), followed
immediately by fixation with paraformaldehyde and permeabilization with Triton X-100;
(C) Cells permeabilized with digitonin and fixed with paraformaldehyde; and (D) Cells
permeabilized with digitonin, incubated with transported buffer for 30 min, prior to
fixation with paraformaldehyde and treatment with Triton X-100. Galectin-3 was

detected with rat anti-galectin-3(250 ug/ml) and fluorescein-conjugated goat anti-rat

immunoglobulin.

60

results suggest that nuclear galectin-3 can be exported during the 30 minute incubation in

TB.

This conclusion is corroborated by Western blot analysis of the galectin-3
polypeptide, both in the nuclear residue as well as in the soluble transported fraction. The
level of galectin-3 in the nuclear residue is decreased considerably after the 30—minute
incubation (Fig. 3, NF). Over the same time period, the level of galectin-3 in the
supernatant rose correspondingly (Fig. 3, TF). Thus, both the immunofluorescence and
immunoblotting analyses yielded consistent data, indicating that there was a rapid export
of the galectin-3 polypeptide from the nuclei of digitonin-permeabilized 3T3 cells. This
export assay has also been carried out using simian Cos-1 cells, with essentially the same

results.

Kinetics of Expgrt and Temmrature Demndence

The rate of export of galectin-3 from digitonin-permeabilized 3T3 cells was
temperature dependent. It was rapid at 37 °C; the half-life of galectin-3 in the nuclear
residue was less than 5 minutes and less than 15% of the original nuclear galectin-3
remained after a 30-minute incubation in TB (Fig. 4). Parallel assays carried out at 25 °C
and 0 °C showed that the rate of exit for galectin-3 decreased with lowering of the
temperature. At 0 °C, for example, at least 40% of the galectin-3 remained with the

nuclear residue over the 30-minute time course monitored. Because of the rapidity of the

61

Figure 3: Western blotting analysis for galectin-3 in the nuclear fraction (NF) and
the transported fraction (T F ) before and after a 30-min. export assay. Galectin-3
was detected using rat anti-galectin-3 (250 ng/ml) and alkaline phosphatase-conjugated
goat anti-rat immunoglobulin. The solid and dotted lines on the right highlight the
positions of migration of the doublet corresponding to galectin-3 in NF at the beginning
of the assay. The galectin-3 in TF consists of only the upper band of the doublet. The
asterisk highlights the presence of a band in TF reactive with anti-galectin—3, most
probably representing a degradation product. The numbers on the left indicate the

positions of migration of molecular weight markers.

 

62

NF TF
t(min) 0 30 0 30-

93-
67-
56"
42-

23—

17-

11-

63

t(min) o 2 4 8 16 30

Figure 4: Temperature dependence of the kinetics of galectin-3 export in the digitonin-
permeabilized cell assay. The amount of galectin-3 remaining in the nucleus was
monitored by Western blot analysis of NF at various times. The solid and dotted lines on the
right highlight the positions of migration of the doublet corresponding to galectin-3. The
arrowhead indicates the position of migration for Sm B, a nuclear protein that is not exported
in the assay, and thus provides an indication that material from approximately the same

number of nuclei were electrophoresed in each lane.

64

export system at 37 °C, all subsequent experiments used samples after an 8-minute

incubation in TB.

In these experiments, the export assay was carried out on cells incubated with
fresh TB, after the removal of the digitonin solution used to permeabilize the plasma
membrane of the cells. We have found that we can obtain essentially the same results by
simply supplementing the digitonin solution with dithiothreitol and protease inhibitors,
without the removal of the detergent. Thus, it appears that the continued presence of
digitonin during the export assay does not extract or deplete factors affecting the transport

itself.

The Exit of Nuclear Galectin-3 is Selective

In the course of these studies, careful analysis Of Western blots indicated that,
consistently, galectin-3 is revealed as a doublet in the nuclear residue whereas only the
upper band of the doublet can be found in the transported fraction (Fig. 3). We had
previously Observed that, in 3T3 fibroblasts, galectin-3 can be found in two isoforms: (a)
a phosphorylated form (pl ~8.2); and (b) a non-phosphorylated species (pl ~8.7). The
former is present both in the nucleus and in the cytoplasm, while the latter is found
exclusively in the nucleus (Cowles et al., 1990). On this basis, the nuclear residue and
the transported fraction were subjected to two-dimensional gel electrophoresis, followed

by Western blotting. The nuclear residue, which contained the galectin-3 doublet on

65
single-dimension SDS-PAGE, yielded two spots on the two-dimensional gels, with the

upper band corresponding to the pl ~8.2 species and the lower band corresponding to the
pI ~8.7 isoform (Fig. 5A). In contrast, the transported fraction, which contained only the

upper band of galectin-3, yielded exclusively the pl ~8.2 spot (Fig. 5B).

On this basis, we conclude that the upper band of the galectin-3 doublet
represented the phosphorylated form while the lower band corresponded to the
nonphosphorylated species. This conclusion is supported by two additional lines of
evidence. First, when the nuclear residue is immunoblotted with anti-phosphoserine,
only the upper band of the galectin-3 doublet showed reactivity (data not shown). This is
consistent with the identification of Ser-6 as the major site of phosphorylation of the
canine homolog of galectin-3 (Huflejt et al., 1993). Second, an export assay was carried
out using 3T3 cells that had been metabolically labeled in the presence of 32P0,. The TF
was subjected to lactose affinity chromatography and the bound material was resolved on
SDS gels. The single galectin-3 band, identified by immunoblotting, was also radioactive

upon autoradiography (data not shown).

Although the exported galectin-3 consists of only the phosphorylated form, the
majority of the galectin-3 molecules was found to be ultimately lost from the nuclear
residue. Moreover, on the basis of the intensity of the immunoblotted bands, the amount
of galectin-3 lost from the nucleus (Fig. 3, NF) can be approximately accounted for by
the amount appearing in the transported fraction (Fig. 3, TF). Thus, it appears that the

non-phosphorylated galectin-3 polypeptide (lower band) is first phosphorylated (to

66

A B NEPHGE {1

1‘3
* *

3.5
(I)

if": - ".53” D
(0

Basic Acidic Basic Acidic

 

Figure 5: Two-dimensional gel electrophoretic analysis for galectin-3 in NF before
the export assay (A) and in the TF after an 8-minute export assay (B). The samples
were subjected to two-dimensional NEPHGE/SDS-PAGE and Western blotting. The
asterisk marks the reference protein rabbit muscle glyceraldehyde 3-phosphate

dehydrogenase, which yielded a pI value of 8.5.

67

become the upper band) prior to export. This notion is supported by a comparison of the
ratios of the upper to lower bands at various time points during transport at 37 °C (Fig. 4).
These results suggest that phosphorylation may be important for the exit of galectin-3

from the nucleus.

The observation that only the phosphorylated form of galectin-3 appears to be
exported suggests the system is highly selective in terms of substrates for export. Thus,
we determined the export or retention of several nuclear proteins, on the basis of antibody
reagents available. Karyopherin-B and Ran, two factors required for the import of NLS-
bearing proteins, were also found to be exported. This conclusion can be derived by
comparing the initial and 8-minute samples of either the nuclear residue (Fig. 6, lanes 1-2
and 5-6) or the exported material (Fig. 6, lanes 3-4 and 7-8). Galectin-l, belonging to the
same lectin family as galectin-3, has also been identified as a nuclear splicing factor
(Vyakamam et al., 1997). Like galectin-3, it is exported from the nucleus in our assay
(Fig. 6, lanes 9-12). On the other hand, polypeptides of the snRNPs, Sm B representing
the core polypeptide of several of the snRNPs and the A polypeptide specific to U1
snRNP, were not exported in our assay (Fig. 6, lanes 13-20). Finally, the hnRNP protein
A1 is exported very slowly (relative to galectin-3), if at all (Fig. 6, lanes 21-24). Thus,
the specificity of the galectin-3 transport system is manifested by: (a) comparing galectin-
3 with other components of the nuclear splicing machinery; and (b) comparing the

isoelectric variants of the galectin-3 polypeptide itself.

68

1234 5678 9101112

1...... --
_——'
O

13 14 15 16 17 18 19 20 21 22 23 24

-

Figure 6: Comparison of several nuclear proteins in terms of their export or retention
in the digitonin-permeabilized cell assay. Samples were subjected to Western blotting
using the following antibodies: lanes 1-4, mouse anti-karyopherin-B (250 ng/ml); lanes 5-8,
mouse anti-Ran (250 ng/ml); lanes 9-12, rabbit anti-galectin-l (1 :200 dilution); lanes 13-16,
human anti-Sm (1: 10,000 dilution); lanes 17-20, human anti-U1 RNP (1: 10,000 dilution);
and lanes 21-24, mouse anti-hnRNP A1 (1: 1000 dilution). The first two lanes of each set
represent NF, at the beginning of the assay and after an 8-minute export period, and the last

two lanes of each set represent the corresponding samples ofTF.

69

Inhibition of Galectin-3 Exmrt

The plant lectin WGA, which binds N-acetyl-D-glucosamine (GlcNAc)-
containing glycoconjugates, has been shown to inhibit the nuclear import of proteins
(Finlay et al., 1987), as well as nuclear export of RNA (Yoneda et al., 1988; Bataille et
al., 1990; Neuman de Vegvar and Dahlberg, 1990; Dargemont and Kuhn, 1992). The
presence of WGA. in our assay inhibited the export of galectin-3 from the nucleus (Fig.
7). This was the case when WGA was added either to the permeabilization buffer
containing digitonin or to TB. The effect of WGA was most pronounced when it was
present during both permeaibilization and transport phases of the assay. Moreover, the
effect of WGA can be reversed by its saccharide ligand, GlcNAc, but not by a non-
binding saccharide, D-galactose (Gal) (Fig. 7). Similar results were obtained when the
export of Ran was analyzed. Inasmuch as it has been shown that WGA has no effect on
diffusion of fluorescent dextrans in and out of the nuclear pores (Jiang and Schindler,
1987; Wolff et al., 1988; Yoneda et al., 1988), these results further argue against the

notion that the observed export of galectin-3 was due to leakage and diffusion.

On the basis of our previous documentation that galectin-3 is required for the
splicing of pre-mRNA (Dagher et al., 1995; Vykarnam et al., 1997) and that its detection
in the nucleus is sensitive to ribonuclease (Laing and Wang, 1988; Hubert et al., 1995),
we tested the effects of the nucleotides m7GpppG and m7GpppA. These correspond to the
monomethylated, inverted guanosine cap structure, m7G(5’)ppp(5’)N, of RNA

polymerase II transcripts and have been implicated to facilitate RNA export from the

70

_ _ WGA + WGA +
WGA GlcNAc Gal

t(min) O 8 8 8 8

GaleCtln‘3 ‘ nun—- - m“ é

m- rams

Ran

Figure 7: The effect of wheat germ agglutinin(WGA)on the export of galectin-3 and
Ran in the digitonin-permeabilized cell assay. Samples of NF, at the beginning of the
assay and after an 8-minute export period, were subjected to Western blotting using
antibodies against galectin-3 (rat anti-galectin-3, 250 ng/ml) and against Ran (mouse anti-
Ran, 250 ng/ml). The concentration of WGA was 500 rig/ml and the concentrations of
GlcNAc and Gal were 100 mM.

71

nucleus (Hamm and Mattaj, 1990). Although neither nucleotide, at a concentration of 2
mM, had an observable effect on galectin-3 export, we were surprised to find that
m7GpppA (2 mM) in combination with the NES peptide (50 ug/ml) yielded partial
inhibition (Fig. 8A). The sequence of the NES peptide corresponds to residues 34-50 of
PKI, found to be responsible for the latter’s export fiom the nucleus, but the addition of
NES peptide (50 rig/ml) by itself had little effect on transport in our assay (data not
shown). Higher concentrations of m7GpppA (4 mM) and NES peptide (700 pig/ml)
yielded greater (but still only partial) inhibition (Fig. 8B). This effect was not observed
with the combination m7GpppG plus NES peptide (Fig. 8A and 8B). Moreover, the
export of Ran was not similarly affected by the combination of m7GpppA plus NES
peptide (Fig. 8B). These results suggest that at least some of the galectin-3 is exported in
association with an RNP and that this RNP contains both the monomethylated cap
structure m7GpppA and polypeptide(s) bearing NES-like sequences. This notion is

schematically illustrated in Figure 9.

72

 

 

 

A_ 2mM 2mM
mGpppG mGpppA
- _ - 50 ug/ml _ 50 ug/ml
. NES NES
t(min) o 8 8 8 8 8
- “—— —— _ #—
-
Galectin-3
B_ 700 11ng
NES
t . _ _ 4mM 4mM
(mm) mGpppG mGpppA
- ‘15—" “‘v - Gal-3
- ~— -. . Ran

Figure 8: The effects of 5'-cap structure nucleotide analogs and of a peptide bearing a
nuclear export signal (NES) on the export of galectin-3 and Ran in the digitonin
permeabilized cell assay. Samples of NF, at the beginning of the assay and after an 8-
minute export period, were subjected to Western blotting using antibodies as described in the
legend to Figure 7. In panel A, the nucleotides, m7GpppG and m7GpppA, were used at a
concentration of 2 mM and the NES peptide was used at a concentration of 50 ug/ml. In
panel B, the corresponding concentrations of nucleotides and peptide were 4 mM and 700

ug/ml, respectively.

73

 

Figure 9: Schematic diagram illustrating the association of galectin-3 with a
ribonucleoprotein complex in TF of the export assay. Galectin-3 is represented by an
oval (Gal3) containing a nuclear export signal (NES). The horizontal line depicts a mRN A,
with a cap structure (mGpppA) at the 5'-end and a poly A tail (dotted line) at the 3'-end. The
sphere (CBP) represents cap-binding protein(s). Other proteins associated with the RNA

may also carry NES sequences.

74
DISCUSSION

Since its initial development by Adam et al. (1990), the digitonin-permeabilized
cell system has contributed much to our understanding of nuclear protein import. By
selectively permeabilizing the plasma membrane, digitonin provides an in vitro system
that appears to preserve much of the native architecture of the cell, allowing for the
elucidation of the components and interactions involved in the transport of
macromolecular substrates across the nuclear envelope. We have now taken advantage of
this selective permeabilization system to analyze the export of galectin-3 from the cell
nucleus. The key findings of our study include: (a) galectin-3 is rapidly and selectively
exported from the nucleus; (b) although both phosphorylated and nonphosphorylated
isoforms of galectin-3 are found in the nucleus, only phosphorylated galectin is identified
in the exported fraction; and (c) while the Sm B and U1 A polypeptides of snRNPs are
retained in the nuclear residue, galectin-l , Ran, and karyopherin-B are also exported in
the same assay; ((1) the rate of export is decreased by cold temperature and by the addition
of WGA; and (e) galectin-3 export is at least partially inhibited by the addition of cap-like

m7GpppA nucleotide plus NES peptide.

At least three types of nuclear export signals (NES) have now been identified: (i)
short sequences enriched in hydrophobic residues, especially leucines (leucine-rich NES),
as originally identified in PKI (F antozzi et al., 1994) and HIV Rev protein (Meyer and

Malirn, 1994); (ii) a 38-amino acid region containing many glycine residues (designated

75
M9), identified on human hnRNP Al (Michael et al, 1995) and on its insect homolog, hrp

36 (Visa et al., 1996); and (iii) a KN S sequence, identified on hnRNP K that, like the M9
region, can function both for import and export (Michael et al, 1997). A sequence
matching the leucine-rich NES, with distinct spacing of the leucine residues, can be found
in galectin-3 homologs from various species. If this sequence indeed serves as an NES
for galectin-3, then one might expect that a peptide corresponding to the NES of PKI
would compete for the export machinery and inhibit the galectin transport. The addition
of the NES peptide failed to affect galectin-3 export. Moreover, in transfection
experiments using a FLAG epitope-tagged galectin-3 construct, we have carried out site-
directed mutagenesis on the leucine residue correponding to Leu-44 of PKI. This leucine
residue had been demonstrated by Wen et al. (1995) to be critical for the export activity
of the NES. We found no difference, however, in the nuclear versus cytoplasmic
distribution of the FLAG-tagged galectin-3 in cells transfected with the wild-type and the

mutant constructs (Y.G. Tsay, unpublished results).

One difficulty in interpreting such experimental results may be related to the
complex nature of many of the molecular species that are exported fi'om the nucleus. It is
generally believed that the RNA substrates for nuclear export are probably all transported
as RNP complexes (see reviews by Izaurralde and Mattaj (1995), by Gerace (1995), and
by Moore (1996)). Most of these RNPs contain multiple polypeptides so it seems
possible that mutational knockout of the putative NES on galectin-3 will not alter its
export properties as long as it remains associated with an RNP that carries other NES-

bearing polypeptide(s) (see Fig. 9). Indeed, this model is supported by the observation

76
that a combination of NES peptide and nucleotide analogs recognized by cap-binding

proteins are required to inhibit galectin-3 export (Fig. 9). Furthermore, this model may
help account for the previous results that the monomethylated cap structure is a
facilitating, but not a required, factor for mRN A export (Harnm and Mattaj, 1990;
Jarrnolowski et al., 1994). The presence of multiple export signals on these RNPS
probably circumvents the absolute requirement of a particular signal during nuclear

export.

Our present results need to be discussed in the context of two other studies that
have used the digitonin-permeabilized cell system to study nuclear export. Moroianu and
Blobel (1995) used digitonin-permeabilized buffalo rat liver cells to preload the nuclei
with fluorescently-labeled human serum albumin covalently conjugated with the NLS of
SV40 large T-antigen. These preloaded nuclei were then incubated to investigate the
conditions for the export of the fluorescent substrate. They found that nuclear export in
their system required Ran and GTP hydrolysis. In contrast, the export of galectin-3 in our
system was not dependent on any exogenously added Ran or GTP. In fact, we found that
Ran was also exported from the nuclei during the course of the transport assay, and , to
the best of our knowledge, this represents the first demonstration on the nuclear export of
Ran. Thus, it seems possible that depletion of Ran (and GTP) during the preloading of
the fluorescently labeled transport substrate and subsequent washing steps may account
for the apparent requirement for these factors, as observed by Moroianu and Blobel

(1995).

77

More recently, Yang et al. (1997) have used digitonin-penneabilized GrH2 rat
hepatoma cells, which express elevated levels of glucocorticoid receptors, to study the
effect of hormone withdrawal on receptor dissociation from chromatin and a novel
nuclear export process that is stimulated by sodium molybdate (Na2M004) and ATP. The
effect of sodium molybdate can be mimicked by protein tyrosine phosphatase inhibitors,
including sodium tungstate, sodium vanadate, and heparin; it is counteracted by inhibitors
of tyrosine kinase such as genistein and tyrphostin AG126 (Yang et al., 1997). The
stimulation of in vitro export by these compounds is not rmique to glucocorticoid
receptors but can also be observed with heat shock proteins hsp70 and hsp90, as well as
hnRNP A1. Several similarities and differences with our present results become

immediately obvious.

First, hnRNP A1 is exported very slowly, if at all, in our 3T3 cell assay. This was
true even in the presence of ATP and sodium molybdate. It should be noted that hnRNP
A1 is known to shuttle between the cytoplasm and the nucleus (Pit‘lol-Roma and
Dreyfuss, 1993) and a nuclear export signal, designated M9, has been identified on the
polypeptide (Michael et al., 1995). The fact that hnRNP A1 failed to be exported rapidly
in our present assay argues strongly against non-specific leakage being responsible for the
observed exit of nuclear components such as galectin-3.. This notion is supported by the
observation that while the nucleus contains both phosphorylated and nonphosphorylated
galectin-3, only the phosphorylated isofonn appears in TF. The export of galectin-3 is

also blocked by WGA, similar to the latter’s inhibition of the transport of preloaded NLS-

78
human serum albumin (Moroianu and Blobel, 1995), and of the ATP- and molybdate-

stimulated export of glucocorticoid receptors (Yang et al., 1997).

Second, the export Of glucocoticoid receptors from the nuclei of GrH2 rat
hepatoma cells required ATP, in addition to molybdate (Yang et al., 1997). On the other
hand, the addition of either ATP or its nonhydrolyzable analog ATPyS had little or no
effect on the export of galectin-3 in our assay. This apparent discrepancy on energy
requirements may reflect the monitoring of different export substrates. ATP hydrolysis
may be required for staging of the glucocorticoid receptors into a subnuclear
compartment for export whereas galectin-3 transport may not need such preparation. In
any case, neither the export of galectin-3 nor the export of glucocorticoid receptors
appear to be sensitive to inhibition by GTPyS, suggesting that GTP hydrolysis was not
required. On the other hand, Moroianu and Blobel (1995) had reported such a GTP

dependence in the export of preloaded NLS-human serum albumin.

Finally, in contrast to the stimulatory effect of molybdate, tungstate, and vanadate
on glucocorticoid receptor export (Yang et al., 1997), we found no effect of these
oxyanions on galectin-3 transport. More strikingly, we observed inhibition of the export
of galectin-3 by vanadyl cations. The basis of the vanadyl versus vanadate difference
may again be related to our idea that at least some of the galectin-3 is exported from

nucleus in the form of an RNP (Fig. 9). Thus, the details of the vanadyl inhibition will be

79

presented in the accompanying manuscript, as a part of the analysis of the chemical

components in the exported fraction (Tsay et al. , accompanying paper).

CHAPTER III
NUCLEAR EXPORT OF GALECTIN-3 IN MOUSE 3T3 FIBROBLASTS

II. CHARACTERIZATION OF THE EXPORTED COMPONENTS

Yeou-Guang Tsay, Nancy Y. Lin, Patricia G. Voss, and John L. Wang

Department of Biochemistry

Michigan State University

East Lansing, MI 48823

80

81

SUMMARY

The nuclear components collected in the transported fiaction of a nuclear export
assay (see accompanying manuscript) was analyzed in terms of the polypeptide and RNA
components. Gel filtration of the exported nuclear material and analysis for galectin-3
showed that the lectin can be found in at least two sets of high molecular weight
complexes (~650 kD and ~60 kD). In the presence of lactose, both of these complexes
are disrupted and galectin-3 chromatographs to a position corresponding to ~ 30 kD
polypeptide. In contrast, the overall elution profile of galectin-1 in the transported
fraction is not drastically altered by lactose. The polypeptide components of the high
molecular weight complexes containing galectin-3 are revealed by affinity adsorption on
a lactose-agarose column, specific elution by lactose, and gel electrophoresis. These
polypeptides are not bound to a control cellobiose column. The transported fraction of
the nuclear export assay also contain RNA. In the low molecular weight range, the RNA
species include tRNA (~ 80 nucleotides) as well as RNAS of ~100, 300, and 650
nucleotides. High molecular weight RNAS, ranging from ~1 kb to 5 kb, include poly A”
mRN A as revealed by hybridization with an oligo(dT) probe. Compounds containing
vanadyl cations (V 02“), such as vanadyl ribonucleoside complex, inhibit the export of
both galectin-3 and RNA in our assay. All of these results are consistent with the notion
that galectin-3 is associated with a ribonucleoprotein complex during its export from the

nucleus.

82
INTRODUCTION

In the previous manuscript, we documented that, in digitonin-penneabilized 3T3
fibroblasts, galectin-3 as well as several other nuclear proteins, is rapidly and selectively
exported from the nucleus (Tsay et al, 1997). Several key findings were made: (a)
although both phosphorylated and nonphosphorylated isoforms of galectin-3 are found in
the nucleus, only phosphorylated galectin-3 is identified in the exported fraction; (b)
while Sm B and U1 A of snRNPs are retained in the nuclear residue, galectin-1, Ran, and
karyopherin-B are also exported in the same assay; and (c) galectin-3 export is inhibited
by the addition of cap-like m7GpppA nucleotide plus a peptide bearing the nuclear export
signal (NES) of the inhibitor of CAMP-dependent protein kinase (PKI). On the basis of
these observations, the possibility was raised that galectin-3 may be exported in
association with a ribonucleoprotein complex (RNP), with many polypeptides bound

directly or indirectly with an RNA species.

One major advantage of the digitonin-permeabilized export assay lies in the fact
that the exported material can be collected in a soluble form, amenable for fiactionation
and biochemical analysis. We have thus carried out preliminary physico-chemical
characterization of the components of the exported nuclear components, with two main
goals: (a) to obtain additional evidence that nuclear galectin-3 is associated with a high

molecular weight complex; and (b) to delineate the number and, if possible, the identity

83

of the polypeptide and RNA species of the galectin-3 complex. The results of these

studies are documented in the present communication.

84

EXPERIMENTAL PROCEDURES

Cell culture and reagents

NIH mouse 3T3 fibroblasts were purchased form the American Type Culture
Collection (Rockville, MD). The cells were grown in Dulbecco's Modified Eagle
Medium containing 10% calf serum, 100 U/ml penicillin, and 100 ug/ml streptomycin at
37°C in a humidified atmosphere of 10% C02. Generally, cells were seeded at a density
of~1.5 x 10‘ cells/cm’, about 30% of the saturation density. Metabolic labeling with 32P-
labeled orthophsophate was carried out at the concentration of 50 uCi/ml culture medium.

Anti-Galectin-3 antibody is a rat monoclonal antibody developed against the Mac-
2 antigen (Ho and Springer, 1982); anti-galectin-l antibody is a polyclonal rabbit
antiserum raised against rat galectin-1 (a generous gift from Dr. Barondes, University of
Califomia, San Francisco); anti-Ran antibody is a mouse monoclonal antibody, purchased
fiom Transduction Laboratories (Lexington, KY).

Vanadyl sulfate hydrate was fiom Aldrich (Milwaukee, WI), vanadyl
ribonucleoside complex (V RC) was from GIBO-BRL (Gaithersburg, MD), and
ribonucleosides were from Boehringer-Mannheim (Indianapolis, IN). Agents purchased
from Sigma (St. Louis, MO) include sodium metavanadate (NaVO3), tungstic acid
(HZW04), sodium molybdate (NaQM004), and all of the secondary antibodies used for

immunofluorescence and immunoblotting analyses.

85

Nuclear exmrt assay

The details of the assay are describe in the accompanying manuscript (Tsay et a1. ,
1997). For transported fraction (TF) used in gel filtration chromatography and
immunoprecipitation analysis, VRC was added at a concentration of 10 mM immediately
following, the collection of TF. TF used for gel filtration and immunoprecipitation was
concentrated approximately 200-fold in Ultrafree-l 5 centrifugal filter device (Millipore,
Bedford, MA) centrifuged at 3,000 x g.

For immunofluorescence experiments, cells seeded on coverslips were incubated
with 40 ug/ml digitonin at 0°C for 5 min. Reagents to be tested for stimulatory or
inhibitory effects then added without removal of the digitonin solution. The cells were
incubated at 25 °C for 8 min. To stop the reaction, cells were incubated in 3.7% of
formaldehyde for 20 min at 0 °C. Fixed cells were washed in phosphate-buffered saline
(PBS) twice for 10 min. Penneabilization with 0.2% Triton X-100 was carried out at
room temperature for 5 min. After two 10—min washes, 0.2% of gelatin was added to the
cells. Staining with antibodies was described in the accompanying manuscript (Tsay et
al., 1997). Concentrations for primary antibodies are 2.5 jig/ml for anti-Ran and 25
ng/Inl for anti-galectin-3. FITC-conjugated anti-mouse IgG and anti-rat IgG antibodies
were both used at 1:250 dilution.

For immunoblotting experiments, cell were treated as described above. After the

incubation at 25°C, the cell residue, designated as the nuclear fiaction (NF), was

86

collected in SDS-PAGE sample buffer. The sample was sonicated for 1 min at room
temperature and heated at 100 °C before loading to a 12.5% polyacrylamide gel. The
polypeptides resolved on polyacrylamide gels were electrotransferred onto Immobilon-P
(Millipore, Bedford, MD) and immunoblotted with the indicated antibodies.
Concentrations of primary antibodies for immunoblotting are 250 ng/ml for anti-Ran, 250
ng/ml for anti-galectin-3 and 1:1000 for anti-galectin-l. All of the secondary antibodies

were used at 1:2000 dilution.

Gel filtration chromatoggphy

About 24 ml of Sephacryl S-300 HR (Phararnacia, Piscataway, NJ) were packed
in a 1 x 30 cm Econo-Column column (Bio-Rad, Hercules, CA) according to the
manufacturer's guide. The column is rinsed with transport buffer (TB) at the rate of 24
ml/hour and then equilibrated with the same buffer at the rate of 8 nil/hour before
application of the sample. Molecular weight standards (100 ug each) include
thyroglobulin (670 kDa), alcohol dehydrogenase (150 kDa) and carbonic anhydrase (29
kDa). Eluted components were collected in 13-drop (~ 0.75 ml) fractions. Protein
concentration in each fraction was determined by the Bradford method (Bradford, 1976).
Typically, TF derived from 4 x 10‘ cells, in a volume of 200 pl, was loaded onto the
column and then eluted with TB. Polypeptide components in each fraction were
precipitated in the presence of 10% trichloroacetic acid and 1% deoxycholate at 4°C for

30 min. The pellet was collected after centrifirgation at 12,000 x g for 15 min. After

87

rinsing with acetone, the pellet was resuspended with 0.1% deoxycholate-0.1 M NaOH
and mixed with 5X SDS sample buffer. The samples were analyzed by SDS-PAGE and

immunoblotting.

Affini adso tion chromato h

For saccharide adsorption experiments, 200 pl of concentrated TF were incubated
with 40 ul of 1:1 slurry of lactose- or cellobiose-agarose (Sigma) overnight at 4°C. These
sample was centrifuged at 12,000 x g for 10 sec, and the supernatant (unbound fiaction)
was removed fi'om the beads. The agarose beads were then washed four times in TB
containing 1% NP-40, 0.5% deoxycholate and 0.1% SDS. After the last wash, bound
materials were eluted with 0.2 M lactose in TB and designated as eluted fiaction. The
eluted fraction was resolved on a 12.5% polyacrylamide gel and analyzed by silver
staining.

For antibody adsorption experiments, concentrated TF was incubated with 3 ug of
anti-galectin-3 overnight at 4°C. Twenty microliters of 1 :1 protein G-agarose were added
to the solution and mixed with rocking at room temperature for 2 hours. The sample was
centrifirged at 12,000 x g for 10 sec, and the supernatant (unbound fraction) was removed

and subjected to RNA analysis.

Analysis of RNA commnents

To samples containing 32P-labeled RNAS were added 0.1% NP40, 0.1% SDS, and

10 ug/ml tRNA and then treated with 10 jig/ml of proteinase K at 37 °C for 10 min.

88
RNAS were then isolated by standard phenol extraction and ethanol precipitation

procedures (Sarnbrook et al., 1989). RNA pellets were resuspended in formaldehyde
loading buffer and resolved on a fonnaldehyde-l .2% agarose gel or an 8% polyacylamide
gel. RNAS were then visualized by autoradiography.

For detection of poly(A)" RNAS, unlabeled RNAS were prepared as described and
then resolved on a formaldehyde-8% agarose gel. RNAS were hybridized in dried gels
with 32P-labeled oligo(dT),2_,8. (GIBCO-BRL, Gaithersburg, MD) at room temperature
using the standard procedure (Sambrook et al., 1989). RNAS hybridized with oligo(dT)

were visualized by autoradiography.

89

RESULTS

Gel Filtration of Exmrted Nuclear Compgnents and Analysis for Galectin-3

In the previous manuscript, we documented the use of digitonin-penneabilized
3T3 fibroblasts to analyze the export of galectin-3 fi'om the nucleus (Tsay, accompanying
manuscript). We defined the material remaining in the nuclear residue alter a given
period of incubation as the Nuclear Fraction (N F). The exported material is collected in
the Transported Fraction (TF). One key observation was that the export of galectin-3 can
be inhibited, at least partially, by the simultaneous addition of the cap-like nucleotide
m7GpppA plus a peptide bearing the NES from PKI. The possibility was raised that
galectin-3 may be exported in association with an RNP, which contain signals that
facilitate nuclear export in both the 5'-cap structure of the RNA species as well as peptide
sequences of component polypeptide(s). Because one major advantage of the digitonin-
penneabilized export assay lies in the fact that the exported material can be collected in a
soluble form amenable for fractionation and biochemical analysis, we have carried out

preliminary physical and chemical characterization of the components of TF.

First, TF was concentrated approximately ZOO-fold and subjected to gel filtration
on a column of Sepharcryl S-300 HR. Individual fiactions eluting from the column were
subjected to SDS-PAGE and Western blotting for galectin-3. The elution profile for

galectin-3 in TF showed three zones, with peaks centered around: (a) ~650 kD; (b) ~60

90

kD; and (c) ~15 kD (Fig. 1A). Previous gel filtration analyses of galectin-3, purified
from 3T3 cells (Roff and Wang, 1983) and from E. coli expression system (Hsu et al.,
1992), showed that the protein chromatographed as a ~30 kD species, corresponding to
the monomeric polypeptide. On this basis, we can immediately conclude that galectin-3
in TF existed in at least two sets of complexes: (a) the ~650 kD group; and (b) the ~60
kD group. The presence of galectin-3 in the zone centered around ~15 kD may
complicate the molecular weight estimates because it suggests that either galectin-3 or an
associated component may interact with the Sephacryl backbone and get retarded in the
chromatographic column. We have carried out similar gel filtration studies using a
number of other matrices (e. g. Sepharose, Sephadex); the same overall conclusion, that

galectin-3 in TF was associated with high molecular weight complexes, was obtained.

The individual fiactions from the Sepharcryl S-300 HR column were also
subjected to Western blotting for galectin-1. There was a broad zone of elution, with a
peak slightly higher in molecular weight and another one just slightly lower in molecular
weight than the 29 kD marker (Fig. 1B). This result is consistent with the report of Cho
and Cummings (1995), who showed that galectin-1, on gel filtration, yielded two poorly
resolved peaks corresponding to the monomeric polypeptide (~15 kD) and a dimer (~30
kD). The ratio of monomer to dimer depended on the concentration of the protein in
solution. Because the members of the galectin family exhibit very similar carbohydrate-
binding specificities, the fact that galectin-l chromatographed at the expected position
suggests that the retardation of galectin-3 in the ~15 kD zone (Fig. 1A) is most probably

not due to protein interactions with the carbohydrate backbone of the Sephacryl matrix.

91

670K 150K 29K
. V V v
A ’ ‘
I
--.—.- - -— ——---- -'“"
B
.2...- .

Figure 1: Gel filtration of the exported nuclear components in TF and analysis for
galectins-l and -3. The column (1 .3 x 29 cm) was packed with Sephacryl S-300 HR and was
equilibrated with TB (containing supplements?) HOW MUCH TF LOADED? IN WHAT
VOLUME? Fractions of 0.5 ml were collected, subjected to SDS-PAGE and
immunoblotting with anti-galectin-3 (panel A) or anti-galectin-I (panel B). The numbers at
the top indicate the positions of migration of molecular weight markers: thyroglobulin

(670 K), alcohol dehydrogenase (150 K), and carbonic anhydrase (29 K).

92

Using purified galectins-l and -3 , previous studies had shown that the
disaccharide ligand lactose has little or no effect on the chromatographic behavior of
these lectins (Roff and Wang, 1993; Hsu et al., 1992; and Cho and Cummings, 1995).
The inclusion of Lac in the chromatographic medium had little or no effect on the overall
position of elution for galectin-1 in TF (Fig. 2). There appeared to be some diminution of
the peak >29 kD and accentuation of the peak < 29 kD. One possibility is that Lac
shifted the monomer-dimer equilibrium in favor of subunit dissociation. An alternative
possibility is that the > 29 kD peak represents some association of galectin-1 with another

component and that this complex is disrupted by Lac binding to galectin-1.

For galectin-3 in TF, however, the inclusion of Lac in the chromatographic
medium drastically altered its elution profile (Fig. 2). The peak at ~650 kD zone
disappeared. The peak at ~60 kD zone was also diminished. The majority of the
galectin-3 chromatographed to a position slightly lower in molecular weight than the 29
kD marker. These results suggest that the high molecular weight complexes with which
galectin-3 is associated in TF is sensitive to Lac-binding. Individual fractions in the ~650
kD zone derived from colmnns carried out in the absence and presence of Lac were
compared by SDS-PAGE and silver staining. The polypeptide compositions of these
fractions fi'om the column with Lac were quite distinct from the corresponding fractions
fi'om columns without the disaccharide (data not shown). These results are consistent
with the notion that Lac removed not only galectin-3 from the ~650 kD zone but also

other polypeptides in the high molecular weight complex. Finally, it should be noted

93

670 K 150 K 29 K

'---—-- -__.---—-

.. -41.-- 1—4' +

Galectin-3
..._. a! ’ -
“a... d +

Galectin-1

Figure 2: The effect of Lac on the gel filtration profile of galectins-1 and -3 in TF. The
column and the conditions of chromatography are as described in the legend to Figure 1. In
experiments testing the effect of Lac, the column was equilibrated and developed with TF

containing 0.2 M Lac.

94

that, even in the presence of Lac, traces of galectin-3 could still be observed over a broad

region of the column, ranging in molecular weight from <600 kD to ~60 kD (Fig. 2).

Affinig Chromatography of Transpgrted Fraction on a Lactose Column

The exported material in TF was also subjected to affinity adsorption on Lac-
agarose beads. After extensive washing, the bound material was eluted with Lac. This
material revealed a very complex composition of polypeptides upon SDS-PAGE and
silver staining (Fig. 3). The most prominent bands corresponded to polypeptides of : (i)
~15 kD; (ii) ~30 kD; (iii) ~40 kD; (iv) ~50 kD; (v) ~58 kD; and (vi) ~93 kD. With the
exception of the ~50 kD band, none of these prominent bands can be found when TF was
subjected to a parallel adsorption on cellobiose-agarose beads (Fig. 3). On the basis of
the Lac-binding property of the galectin family members and on the basis of
immunoblotting, galectins-l and -3 can be identified. The other bands most probably
represent polypeptides of a complex that have bound to the Lac beads through their
association with either galectins-l or -3. The complexity of the polypeptide profile
precludes a detailed comparison. However, it should be noted that the pattern presented
in Figure 3 (Lac affinity adsorption) is very similar to: (a) that obtained from fractions of
the ~650 kD zone of the Sephacryl S-300 HR column; and (b) that observed when
nucleoplasm Of 3T3 fibroblasts is fractionated on an N-(e-aminocaproyl)-D-

galactosarnine-Sepharose column and eluted with galactose (Laing and Wang, 1988).

95

Figure 3: Affinity adsorption of the polypeptide components in TF on Lac-
agarose and cellobiose-agarose. Parallel samples of 200 111 were incubated with 40 pl
of Lac-agarose or cellobiose-agarose. The material bound to the beads were eluted with
TB containing 0.2 M Lac, subjected to SDS-PAGE and silver staining. The numbers at

the left indicate the positions of migration of molecular weight markers.

96

221—
133..

93—
67—

56—

42-

28—

 

23- '

17—

11-

(kDa)

97

The gel filtration studies on TF in the presence and absence of lactose suggested
that the binding of the disaccharide disrupted the high molecular weight complexes
containing galectin-3 (Fig. 2). Therefore, the question may be raised as to how galectin-
3-associated polypeptides can still be isolated on the basis of Lac-affinity adsorption (Fig.
3). One possible explanation may lie in our previous parenthetical note that, even in the
presence of Lac, traces of galectin-3 could still be Observed over a broad range of the gel
filtration column, from <600 kD to ~60 kD (Fig. 2). This may reflect dynamic and
continuous reequilibration of galectin-3 with the high molecular weight complex and it is
on this basis that the Lac affinity beads can trap some of the associated polypeptides. It
may also be worthy to note that the efficiency of galectin-3 isolation (amount of galectin-
3 in the bound and Lac-eluted fiaction as a percent of galectin-3 in the unfractionated
material) fi'om TF was much lower than the efficiency of adsorption of purified
recombinant galectin-3 onto the same Lac-agarose matrix. One possible implication of
this observation is that the affinity of galectin-3 for its saccharide ligand is decreased in
complex formation. Alternatively, galectin-3 in TF may already be occupied with

carbohydrate ligands with which Lac-agarose must compete for binding.

Effect of Vanadyl Cations on the Exmrt of Galectin-3

The inhibition of galectin-3 export by the combination of 5’ cap-like nucleotide
and NES peptide had suggested the possibility that the lectin was co-exported with
RNA(s) (Tsay et al., 1997). The present finding that a significant portion of the exported

galectin-3 can be found in a high molecular weight complex lends support to the

98

hypothesis. On the basis of previous reports documenting that VRC can stabilize RNA
species from nuclease degradation (Berger and Birkenmeier, 197 9), we tested the effect
of VRC on the preservation of the galectin-3 high molecular weight complex in TF. To
our surprise, we found that VRC (10 mM) was a potent inhibitor of nuclear export in our
assay (Fig. 4A). The active ingredient appears to be vanadyl (V O”) cations, inasmuch as
vanadyl sulfate (10 mM) also inhibited nuclear export but little or no inhibition was
Observed with a mixture of the four ribonucleosides (10 mM). The oxoanion of
vanadium, vanadate (V 03'), did not Show inhibition at the same concentration. Nor did

other transition metal oxoanions, molybdate and tungstate (data not shown).

A comparison of the dose-response curves of inhibition showed that VRC was
potent at concentrations as low as 5 mM (Fig. 48) while strong inhibition was observed
only with higher concentrations (10 mM or greater) of vanadyl sulfate (Fig. 4C). The
conclusion that VRC was more effective than vanadyl sulfate in suppressing nuclear
export can also be derived from immunofluorescence analysis for galectin-3 in the
nuclear residue. The level of galectin-3 appeared to be higher in the nuclear residue of
VRC (10 mM)-treated cells (Fig. 5D) than the corresponding cells treated with vanadyl
sulfate (10 mM) (Fig. 5E). Surprisingly, the simultaneous application of VRC and
vanadyl sulfate (10 mM each) resulted in even greater inhibition than either reagent alone
(Fig. 5F). This suggests the possibility that the effects of the two reagents may be
additive and thus may be acting via different pathways or different populations of

molecules. Parenthetically, the inhibitory effect of vanadyl cations is more potent that

99

 

Figure 4: The effect of vanadyl cations (V01) on the nuclear export of galectin-3,
assayed by immunoblotting NF. In panels A-C, the first left hand lane represents NF at the
beginning of the export assay. All the other lanes in each panel represent NF at the end of an

8-minute export period. Panel A: -, no addition; VRC, vanadyl ribonucleoside complex (10

mM); V0”, vanadyl sulfate (10 mM); Ribonucleoside, mixture of four ribonucleosides (2.5
mM each). Panels B and C: dose-response curves for VRC and Vanadyl sulfate,

respectively.

 

Figure 5: The effect of vanadyl cations (VO“) on the nuclear export of galectin-3,
assayed by immunofluorescence analysis of NF. Galectin-3 was detected using rat anti-
galectin-3 (250 ng.ml) and fluorescein-conjugated goat anti-rat immunoglobulin. A: NF at
the beginning of the export assay. B-F: NF at the end of an 8-minute export period. B: no
addition; C: wheat germ agglutinin (500 ug/ml); D: VRC (10 mM); E: vanadyl sulfate (10
mM); and F: VRC and vanadyl sulfate (10 mM each).

101

that of wheat germ agglutinin (Fig. 5C), a conclusion also corroborated by subjecting NF

to Western blot analyses.

Finally, the inhibitory effect of vanadyl cations on nuclear export is not restricted
to galectin-3; a similar set of results was also obtained by analyzing for Ran.
Immunofluorescence analysis of NF at the beginning of the export assay (Fig. 6A) and
after an 8-minute export period (Fig. 6B) showed that Ran was rapidly exported. This is
in agreement with the immunoblotting data presented in the accompanying manuscript
(Tsay et al., 1997). To the best of our knowledge, the immunoblotting and
immunofluorescence data together represent the first rigorous documentation that Ran is
exported from the nucleus. Like that for galectin-3, the export of Ran was sensitive to
inhibition by wheat germ agglutinin (Fig. 6C), by VRC (Fig. 6 D), and by vanadyl sulfate
(Fig. 6E), with the strongest effect observed with a combination of VRC and vanadyl

sulfate (Fig. 6F).

Analysis of RNA Commnents in the Transpgrted Fraction

An export assay was carried out using 3T3 cells that had been metabolically
labeled with [32P]orthophosphate. The RNA components in TF collected from this assay
were extracted, precipitated, and resolved on separate polyacrylamide and agarose gels.
Autoradiography of the polyacrylamide gels revealed low molecular weight RNAS, with
bands corresponding to: (i) ~80 nucleotides; (ii) ~100 nucleotides; (iii) ~300 nucleotides;

(iv) ~650 nucleotides; (v) ~800 nucleotides; and (vi) >1000 nucleotides (Fig. 7A). With

102

 

Figure 6: The effect of vandyl cations NO“) on the nuclear export of Ran, assayed by
immunofluorescence analysis of NF. Ran was detected using mouse anti-Ran (250 ng/ml)
and fluorescein-conjugated goat anti-mouse immunoglobulin. A: NF at the beginning of the
export assay. B-F: NF at the end of an 8-rninute export period. B: no addition; C: wheat
germ agglutinin (500 ug/ml); D: VRC (10 mM); E: vanadyl sulfate (10 mM); and F: VRC

and vanadyl sulfate (10 mM each).

103

Figure 7: Gel electrophoretic analysis of the RNA components in TE when the
export assay is carried out in the absence and presence of vanadyl sulfate. The
export assay was carried out using 3T3 cells metabolically labeled with
[32P]orthophosphate and the RNA components of TF were isolated and resolved on an 8%
polyacrylamide gel (panel A) or a 1.2% agarose gel (panel B). The radioactive RNA
components were revealed by autoradiography. The numbers on the left indicate
positions of migration of size markers in nucleotides (panel A) or in kb (panel B). In
experiments testing the effect of vanadyl cations, the export assay was carried out in the

presence of 10 mM vanadyl sulfate.

1.35-
0.87-

0.60-

0.31-
0.23-

0.19-

0.12-

0.07-

(kb)

 

 

VO2+

104

4.9-
3.5-

2.0-

1.6- .

1.1-
0.6-

(kb)

 

105
the exception of tRNA (75-85 nucleotides), the identities of the other small RNAS are not

known. Autoradiography of the agarose gel revealed a smear ranging from 0.7 kb to ~5
kb, with prominent bands at ~ 1.2 kb and ~ 3 kb (Fig. 7B). An Oligo dT probe was used
to hybridize to the high molecular weight RNA; on this basis, RNA species ranging fiom
1 kb to ~5kb can be identified as poly(A)+ mRN A (data not shown). Therefore, it appears
that both low molecular weight RNA, as well as mRNA, are exported in our assay and

can be collected in TF.

We have made several attempts to delineate the RNA species associated with the
high molecular weight galectin-3 complex, using either Lac affinity adsorption or
immunoprecipitation. Due to technical difficulties that we have yet overcome, however,
we have been unsuccessful in isolating and identifying the RNA components of this
complex. By comparing the level of radioactivity and the RNA species in the u_nbgu_n_d_
fractions the Lac-beads or anti-galectin-3-beads with the corresponding unbound fractions
of control beads (e. g. cellobiose-beads), we have ascertained that radioactive RNA is
indeed bound to those beads that should trap galectin-3. Thus, what remains is the

delineation of the number and identity of the RNA species.

Just as we had demonstrated for the export of galectin-3 (and Ran) in terms of
polypeptides, vanadyl cations also inhibited the export of RNA in our assay. The
transport of both low molecular weight RNA species (Fig. 7A), as well as high molecular

weight mRNA (Fig. 73), were suppressed.

106

DISCUSSION

Taking advantage of the fact that the material in TF of the digitonin-permeabilized
nuclear export assay can be collected in a soluble form, we have carried out initial
characterization of the components of TF. The key findings of the present study include:
(a) gel filtration of TF and analysis for galectin-3 showed that the lectin is associated with
at least two high molecular weight complexes (~650 kD and ~60 kD); (b) both complexes
are sensitive to disruption by the presence of the galectin ligand, Lac; (c) the polypeptide
composition of the high molecular weight galectin-3 complex appears to be very
complicated, with prominent bands of about 15 kD, 30 kD, 40 kD, 50 kD, 58 kD, and 90
kD; (d) TF also contains RNA, including small nuclear RNAS, as well as poly A+ mRNA;

and (e) vanadyl cations (V 02“) inhibit the export of galectin-3 and the RNAS.

These results provide new levels of information to two important previous
observations. First, galectin-3 has been identified as a required factor in the splicing of
pre-mRNA, assayed in a cell-free system (Dagher et al., 1995; Vyakamam et al, 1997).
Saccharides such as Lac that bind to galectin-3 with high affinity inhibit the splicing
reaction. The present finding that the high molecular weight galectin-3 complex is
disrupted by Lac may provide a mechanism for the effect of the saccharide in the splicing
reaction. In both the inhibition of splicing and in the disruption of the high molecular

weight galectin-3 complex, it still remains to be determined whether Lac is competing

107

with a glycoconjugate ligand for the carbohydrate-binding site of galectin-3 or it is
inducing a conformational change upon binding to the galectin-3 polypeptide. Using
differential scanning calorimetry, it has in fact been demonstrated that Lac binding to
galectin-3 induces a conformation change in the polypeptide (Agrwal et al., 1993).
Coupled with our failure, despite various attempts, to identify a glycoconjugate ligand for
galectin-3 in the nuclear fraction, we tend to favor the notion that a conformational
change in the galectin polypeptide upon Lac binding dissociates the protein from its

complex. This may, in turn, lead to loss of splicing activity.

Second, we had documented, in the accompanying manuscript, that the nuclear
export of galectin-3 was sensitive to inhibition by a combination of a nucleotide
mimicking the 5’ cap structure of mRNAs and of a peptide bearing the NES from PKI.
Coupled with its required role in the splicing reaction, the results suggested that galectin-
3 may be exported in the form of an RNP, with many polypeptides bound directly or
indirectly with an RNA species. Thus, signals that facilitate nuclear export in both the 5’
cap structure of the RNA species and the peptide sequences of the component
polypeptides needed to be neutralized before inhibition of export could be observed.
Indeed, we found in the present study that gel filtration of the exported material yielded
galectin-3 in at least two sets of high molecular weight complexes. It should be noted
that although Lac disrupts the galectin-3 complex(es), the saccharide actually has little or
no effect on the rate of the protein’s export from the nucleus (Y.G. Tsay, unpublished
observations). Even in the presence of Lac, however, traces of galectin-3 could be

observed over a broad range of the gel filtration column (from <600 kD to ~60 kD) and

108

these putative complexes may continue to serve the role of providing multiple signals for

nuclear export.

Although we have obtained some information on the complexity of the
polypeptide composition of the galectin-3-associated particle and although we can
document that TF contains RNA, we have yet to delineate the number and identity of the
RNA species. If galectin-3 is indeed exported in the form of an RNP, then it would join
two other pre-mRNA splicing factors that accompany their RNA substrates out of the
nucleus. Both hnRNP A1 (Mayeda et al., 1992; Caceres et al., 1994) and the cap-binding
complex that recognize the monomethylated guanosine cap structures at the 5’-end of
RNA polymerase II transcripts (Izaurralde et al., 1994; Lewis et al., 1996) travel out of
the nucleus with mRNAs (Visa et al., 1996a and Visa et al., 1996b). Once out of the
nucleus, whether the RNA stays with galectin-3, as is with hnRNP A1 (Visa et al.,
1996b), or the splicing factor is dissociated from the RNP, as is with the cap-binding

complex (Visa et al., 1996a), are issues that need to be studied.

In the course of the present study, we found that vanadyl cations strongly
inhibited the nuclear export of both galectin-3 and RNAS. However, no effect could be
demonstrated with vanadate, molybdate, or tungstate. This is to be contrasted to the
stimulatory effect of these oxyanions on the nuclear export of the glucocorticoid receptor
(Yang et al., 1997). Vanadium exhibits complex chemistry, fluctuating between vanadyl
(oxidation state +4) and vanadate (oxidation state +5), with possible reduction-oxidation

interconversions (Macara et al., 1980; Li et al., 1996). Over the short time course of our

109

export assays, however, the inhibitory effect on galectin-3 export appears to be restricted
to compounds that give rise to vanadyl cations (V RC, vanadyl sulfate) whereas no such
effects were observed with vanadates. This is consistent with recent observations that
while both vanadyl cations and vanadate oxyanions can target protein phosphotyrosine
phosphatases, distinct vanadate-independent and vanadyl-dependent pathways can be

dissected (Li et al., 1996).

Beside protein phosphotyrosine phosphatases, another set of targets susceptible to
inhibition by ribonucleoside-vanadyl complexes are nucleases (Berger and Birkenmeier,
1979). Thus, it seems possible that VRC indirectly affects galectin-3 export through
inhibition of a ribonuclease(s). This enzyme may have specifically digested an RNA that
anchors galectin-3 -containing RNPS to the nuclear structure. Galectin-3 export is not
allowed, when the anchor remains intact due the inhibition of the nuclease. This notion is
consistent with our previous observation that association of galectin-3 with nuclear
structure was sensitive to treatment with ribonuclease (Laing and Wang, 1988; Wang et
al., 1995). Since VRC has been shown to have similar inhibitory effects on nuclear
export of Ran and galectin-l, if this peculiar mechanism does exist, it may also regulate
the export of Ran and galectin-1. So far, crucial evidence is still lacking and this

possibility awaits firrther investigation.

110

CONCLUSIONS

In this manuscript, we demonstrate that galectin-3 has nuclear shuttling activity;
i.e., it can move between the nucleus and cytoplasm bidirectionally. The .
nucleocytoplasmic distribution of galectin-3 is , in fact, a net result of counteraction
between nuclear import and export. When nuclear import is predominant, this will lead
to nuclear localization. In contrast, when nuclear export dominates over import,
cytoplasmic localization will be seen.

As a pre-mRN A splicing factor, galectin-3 shows differentiation localization as a
function of proliferative states of cells, which distinguishes galectin-3 from other splicing
factors. The major implication is that the splicing process may be modulated by nuclear
accumulation of galectin-3. In conjunction with the fact that galectin-3 can shuttle
between the nucleus and cytoplasm, it is possible that nuclear import may serve as a
mechanism that upgrades RNA splicing activities, and nuclear export is a mechanism that
down-regulates RNA splicing activities.

On the other hand, we also report that galectin-3 appears to be exported in
complexes that contain RNAS capped with mGpppA dinucleotide and polypeptides
bearing NES signals. This finding appears to suggest that galectin-3, along with other
components in these complexes, may participate in RNA metabolism or other cellular

processes subsequently. In other words, nuclear export of galectin-3 is not only a way to

111

turn ofir nuclear activities, but also a mechanism to turn on certain cytoplasmic activities.
It will be interesting to study the fate of these ribonucleoprotein complexes in cytoplasm

as well to explore the roles of galectin-3 in the process.

LIST OF REFERENCES

112

LIST OF REFERENCES

Ackerman, S. J., S. E. Corrette, H. F. Rosenberg, J. C. Bennett, D. M. Mastrianni, A.
Nicholson Weller, P. F. Weller, D. T. Chin, and D. G. Tenen. 1993. Molecular cloning
and characterization of human eosinophil Charcot-Leyden crystal protein
(lysophospholipase). Similarities to IgE binding proteins and the S-type animal lectin
superfamily. J. Immunol. 150:456-468.

Adam, S. A., R. S. Marr, and L. Gerace. 1990. Nuclear protein import in permeabilized
mammalian cells requires soluble cytoplasmic factors. J. Cell Biol. 111:807-816.

Agrwal, N., Q. Sun, S.-Y. Wang, and J. L. Wang. 1993. Carbohydrate-binding protein 35:
I. Properties of the recombinant polypeptide and the individuality of the domains. J. Biol.
Chem. 268:14932-14939.

Ahmed, H. and G. R. Vasta. 1994. Galectins: conservation of functionally and
structurally relevant amino acid residues defines two types of carbohydrate recognition
domains. Glycobiology 4: 545- 549.

Aitchison, J. D., G. Blobel, and M. P. Rout. 1996. Kap104p: a karyopherin involved in
the nuclear transport of messenger RNA binding proteins. Science 274:624-627.

Akey, C. W. and M. Raderrnacher. 1993. Architecture of the Xenopus nuclear pore
complex revealed by three dimensional cryo-electron microscopy. J. Cell Biol. 122:1-19.

Alzhanova-Ericsson, A., X. Sun, N. Visa, E. Kiseleva, T. Wurtz, and B. Daneholt. 1996.
A protein of the SR family of splicing factors bind extensively to exonic Balbiani ring

pre-mRN A and accompanies the RNA fiom the gene to the nuclear pore. Genes & Devel.
10:2881-2893.

Baeuerle, P. A., and D. Baltimore. 1996. NF-sz ten years after. Cell 87:13-20.

ll3

Barondes, S. H. 1997. Galectins: A personal overview. Trends in Glycoscience and
Glycotechnology 9: 1-7.

Barondes, S. H., D. N. W. Cooper, M. A. Gitt, and H. Leffler. 1994. Galectins. J. Biol.
Chem. 269:20807-20810.

Bastos, R., L. R. de Pouplana, M. Enarson, K. Bodoor, and B. Burke. 1997. Nup84, a
novel nucleoporin that is associated with CAN/Nup214 on the cytoplasmic face of the
nuclear pore complex. J. Cell Biol. 137:989-1000.

Bastos, R., A. Lin, M. Enarson, and B. Burke. 1996. Targeting and function in mRNA
export of nuclear pore complex protein Nup153. J. Cell Biol. 134:1141-1156.

Bataille, N., T. Helser, and H. M. Fried. 1990. Cytoplasmic transport of ribosomal
subunits microinjected into the Xenopus laevis oocyte nucleus: a generalized, facilitated
process. J. Cell Biol. 111:1571-1582.

Belanger, K. D., M. A. Kenna, S. Wei, and L. I. Davis. 1994. Genetic and physical
interactions between Srplp and nuclear pore complex proteins Nuplp and Nup2p. J. Cell
Biol. 126:619-630.

Bennett, M., S. Michaud, J. Kingston, and R. Reed. 1992. Protein components
specifically associated with prespliceosome and spliceosome complexes. Genes & Devel.
6: 1986-2000.

Bennett, M. and R. Reed. 1993. Correspondence between a mammalian spliceosome
component and an essential yeast splicing factor. Science 262: 105-108.

Berger, S. L. and C. S. Birkenmeier. 1979. Inhibition of intractable nucleases with
ribonucleoside-vanadyl complexes: isolation of messenger ribonucleic acid from resting
lymphocytes. Biochemistry 18:5 143-5 149.

Bischoff, F. R., H. Krebber, T. Kempf, I. Hermes, and H. Ponstingl. 1995. Human
RanGTPase-activation protein RanGAPl is a homologue of yeast Rnalp involved in
mRNA processing and transport. Proc. Natl. Acad. Sci. USA 92:1749-1753.

Bischoff, F. R. and H. Ponstingl. 1991. Mitotic regulator protein RCCl is complexed
with a nuclear Ras-related polypeptide. Proc. Natl. Acad. Sci. USA 88:10830-10834.

114

Boer, J. M., J. M. A. Van Deursen, H. J. Croes, J. A. M. Fransen, and G. C. Grosveld.
1997. The nucleoporin CAN/Nup214 binds to both the cytoplasmic and the

nucleoplasmic sides of the nuclear pore complex in overexpressing cells. Exp. Cell Res.
232:182-185.

Bogerd, A. M., J. A. Hoffman, D. C. .Amberg, G. R. Fink, and L. 1. Davis. 1994. nupl
mutants exhibit pleiotropic defects in nuclear pore complex function. J. Cell Biol.
127:319—332.

Bogerd, H. P., R. A. Fridell, R. E. Benson, J. Hua, and B. R. Cullen. 1996. Protein
sequence requirements for function of the human T-cell leukemia virus type 1 Rex

nuclear export signal delineated by a novel in vivo randomization-selection assay. Mal.
Cel. Biol. 16:4207-4214.

Bogerd, H. P., R. A. Fridell, S. Madore, and B. R. Cullen. 1995. Identification of a novel
cellular cofactor for the Rev/Rex class of retroviral regulatory proteins. Cell 82:485-494.

Bonifaci, N., J. Moroianu, A. Radu, and G. Blobel. 1997. Karyopherin b2 mediates
nuclear import of a mRN A binding protein. Proc. Natl. Acad. Sci. USA 94:5055-5060.

Borer, R. A., C. F. Lehner, h. M. Eppenberger, and E. A. Nigg. 1989. Major nucleolar
proteins shuttle between nucleus and cytoplasm. Cell 56:379-390.

Bradford, M. M. 1976. A rapid and sensitive method for the quantitation of microgram
quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem.
72:248-254.

Buss, F. and M. Stewart. 1995. Macromolecular interactions in the nucleoporin p62
complex of rat nucler pores: binding of nucleoporin p54 to the rod domain of p62. .1. Cell
Biol. 128:251-261.

Byrd, D. A., D. J. Sweet, N. Pante, K. N. Konstantinov, T. Guan, A. C. S. Saphire, P. J.
Mitchell, C. Cooper,S., U. Aebi, and L. Gerace. 1994. Tpr, a large coiled-coil protein
whose amino-terminus is involved in activation of oncogenic kinases, is localizaed to the
cytoplasmic surface of the nuclear pore complex. J. Cell Biol. 127: 1515-1526.

Caceres, J. F., S. Stamm, D. M. Helfrnan, and A. R. Krainer. 1994. Regulation of

alternative splicing in vivo by overexpression of antagonistic splicing factors. Science
265:1706-1709.

115

Chandran, U. R. and D. B. DeFranco. 1992. Intemuclear migration of chicken
progesterone receptor, but not simian virus-40 large tumor antigen, in transient
heterokaryons. Mol. Endocrinol. 6:837-844.

Cherayil, B. J., S. J. Weiner, and S. Pillai. 1989. The Mac-2 antigen is a galactose-
specific lectin that binds IgE. .1. Exp. Med. 170:1959-1972.

Cho, M. and R. D. Cummings. 1995. Galectin-1, a b-galactoside-binding lectin in
Chinese hamster ovary cells: 1. Physical and chemical characterization. J. Biol. Chem.
270:5198-5206.

Cleves, A. E., D. N. W. Cooper, S. H. Barondes, and R. B. Kelly. 1996. A new pathway
for protein export in Saccharomyces cerevisiae. J. Cell Biol. 133:1017-1026.

Cooper, D. N., S. M. Massa, and S. H. Barondes. 1991. Endogenous muscle lectin
inhibits myoblast adhesion to larninin. J. Cell Biol. 115: 1437-1438.

Cooper, D. N. W. and S. H. Barondes. 1990. Evidence for export of a muscle lectin from
cytosol to extracellular matrix and for a novel secretory mechanism. J. Cell Biol.
110:1681-1691.

Cooper, D. N. W., B. P. Boulianne, S. Charlton, E. M. Farrell, A. Sucher, and B. C. Lu.
1997. Fungal galectins, sequence and specificity of two isolectins from Coprinus
cinereus. J. Biol. Chem. 272:1514-1521.

Cowles, E. A., N. Agrwal, R. L. Anderson, and J. L. Wang. 1990. Carbohydrate-binding

protein 35. Isoelectric points of the polypeptide and a phosphorylated derivative. J. Biol.
Chem. 265:17706-17712.

Cowles, E. A., I. K. Moutsatsos, J. L. Wang, and R. L. Anderson. 1989. Expression of
carbohydrate binding protein 35 in human fibroblasts: comparisons between cells with
different proliferative capacities. Exp. Gerontol. 24:577-585.

Dagher, S., J. L. Wang, and R. J. Patterson. 1995. Identification of galectin-3 as a factor
in pre-mRNA splicing. Proc. Natl. Acad. Sci. USA 92:1213-1217.

Daneholt, B. 1997. A look at messenger RNP moving through the nuclear pore. Cell
88:585-588.

116

Dargemont, C. and L. C. Ktihn. 1992. Export of mRNA from microinjected nuclei of
Xenopus laevis oocyte. J. Cell Biol. 118:1-9.

Davis, L. I. 1995. The nuclear pore complex. Annu. Rev. Biochem. 64:865-896.

Davis, L. I. and G. Blobel. 1987. Nuclear pore complex contains a family of
glycoproteins that includes p62: Glycosylation through a previously unidentified cellular
pathway. Proc. Natl. Acad. Sci. USA 84:7552-7556.

Dreyfuss, G., M. J. Matunis, S. Pii‘lol-Roma, and C. G. Burd. 1993. hnRNP proteins and
the biogenesis of mRNA. Annu. Rev. Biochem. 62:289-321.

Emig, S., D. Schmalz, M. Shakibaei, and K. Buchner. 1995. The nuclear pore complex
protein p62 is one of several sialic acid-containing proteins of the nuclear envelope. J.
Biol. Chem. 270:13787-13792.

Fantozzi, D. A., A. T. Harootunian, W. Wen, S. S. Taylor, J. R. Ferarnisco, R. Y. Tsien,
and J. L. Meinkoth. 1994. Therrnostable inhibitor of cAMP-dependent protein kinase
enhances the rate of export of the kinase catalytic subunit from the nucleus. J. Biol.
Chem. 269:2676-2686.

Finlay, D. R., D. D. Newmeyer, T. M. Price, and D. J. Forbes. 1987. Inhibition of in vitro
nuclear transport by a lectin that binds to nuclear pores. J. Cell Biol. 104:189-200.

Fischer, U., J. Huber, W. C. Boelens, I. W. Mattaj, and R. Lithrrnann. 1995. The HIV-l
Rev activation domain is a nuclear export signal that accesses an export pathway used by
specific cellular RNAS. Cell 82:475-483.

F lach, J., M. Bossie, J. Vogel, A. Corbett, T. Jinks, D. A. Willins, and P. A. Silver. 1994.
A yeast RNA-binding protein shuttles between the nucleus and the cytoplasm. Mol. Cell
Biol. 14:8399-8407.

Fomerod, M., J. van Deursen, S. van Ball, A. Reynolds, .D. Davis, K. G. Murti, J.
Fransen, and G. Grosveld. 1997. The human homologue of yeast CRMl is in a dynamic

subcomplex with CAN/Nup214 and a novel nuclear pore component Nup88. EMBO J.
16:807-816.

117

Fridell, R. A., R. Truant, L. Thorne, R. E. Benson, and B. R. Cullen. 1997. Nuclear
import of hnRNP A1 is mediated by a novel cellular cofactor related to karyopherin-b. J.
Cell Sci. 110:1325-1331.

Fritz, C. C., M. L. Zapp, and M. R. Green. 1995. A human nucleoporin-like protein that
specifically interacts with HIV Rev. Nature 376:530-533.

Fukuda, M., I. Gotoh, Y. Goto, and E. Nishida. 1996. Cytoplasmic localization of
motigen-activated protein kinase kinase directed by its NHz-terminal, leucine-rich short

amino acid sequence, which acts as a nuclear export signal. J. Biol. Chem. 271 :20024-
20028.

Gerace, L. 1995. Nuclear export signals and the fast track to the cytoplasm. Cell 82:341-
344.

Gerace, L., Y. Ottaviano, and C. Kondor-Koch. 1982. Identification of a major
polypeptide Of the nuclear pore complex. J. Cell Biol. 95:826-837.

Gitt, M.A., S. M. Massa, H. Leffler; and S. H. Barondes. 1992. Isolation and expression
of a gene encoding L-14-II, a new human soluble lactose-binding lectin. J. Biol. Chem
267:10601-10606

Gbrlich, D., R. Kraft, S. Kostka, F. Vogel, E. Hartrnann, R. A. Laskey, I. W. Mattaj, and
E. Izaurralde. 1996. Importin provides a link between nuclear import and U snRNA
export. Cell 84:21-32.

GOrlich, D. and I. W. Mattaj. 1996. Nucleocytoplasmic transport. Science 271:1513-1518.

GOrlich, D., S. Prehn, R. A. Laskey, and E. Hartrnann. 1994. Isolation of a protein that is
essential for the first step of nuclear protein import. Cell 79:767-778.

Greber, U. F. ., A. Senior, and L. Gerace. 1990. A major glycoprotein of the nuclear
pore complex is a membrane-spanning polypeptide with a large lumenal domain and a
small cytoplasnric tail. EMBO J. 9:1495-1502.

Guiochon-Mantel, A., P. Lescop, S. Christin-Maitre, H. Loosfelt, M. Perrot-Applanat,

and E. Milgrom. 1991. Nucleocytoplasmic shuttling of the progesterone receptor. EMBO
J. 10:3851-3859.

118

Hallberg, E., R. W. Wozniak, and G. Blobel. 1993. An integral membrane protein of the
pore membrane domain of the nuclear envelope contains a nucleoporin-like region. J.
Cell Biol. 122:513-522.

Hamann, K. K., E. A. Cowles, J. L. Wang, and R. L. Anderson. 1991. Expression of
carbohydrate binding protein 35 in human fibroblasts: Variations in the levels of mRNA,

protein, and isoelectric species as a function of replicative competence. Exp. Cell Res.
196:82-91.

Hamm, J. and I. W. Mattaj. 1990. Monomethylated cap structures facilitate RNA export
form the nucleus. Cell 63:109-118.

Hanada, M., C. Aimé-Sempé, T. Sato, and J. C. Reed. 1995. Structure-function analysis
of Bel-2 protein. Identification of conserved domains important for homodimerization
with Bcl-2 and heterodirnerization with Bax. J. Biol. Chem. 270:11962-11969.

Hart, G. W., G. D. Holt, and R. S. Haltiwanger. 1988. Nuclear and cytoplasmic
glycosylation: novel saccharide linkages in unexpected places. T IBS 13:3 80-3 84.

Hirabayashi, J ., M. Satoh, and K.-i. Kasai. 1992. Evidence that Caenorhabditis elegans
32-kDa b-galactoside-binding protein is homologous to vertebrate b—galactoside-binding
lectins. J. Biol. Chem. 267:15485-15490.

Ho, M. K. and T. A. Springer. 1982. Mac-2, a novel 32,000 Mr mouse macrophage

subpopulation-specific antigen defined by monoclonal antibodies. J. Immunol. 128: 1221-
1228.

Hong, W., M. Bennett, Y. Xiao, R. F. Kramer, C. Wang, and R. Reed. 1997. Association
of U2 snRNP with the spliceosomal complex B. Nucleic Acids Res 25:354-361.

Hopper, A. K., H. M. Traglia, and R. W. Dunst. 1990. The yeast RNA] gene product
necessary for RNA processing is located in the cytosol and apparently excluded from the
nucleus. J. Cell Biol. 111:309-321.

Hubert, M., S. Y. Wang, J. L. Wang, A. P. Seve, and J. Hubert. 1995. Intranuclear
distribution of galectin-3 in mouse 3T3 fibroblasts: Comparative analyses by
immunofluorescence and immunoelectron microscopy. Exp. Cell Res. 220:397-406.

119

Huflejt, M. E., C. W. Turck, R. Lindstedt, S. H. Barondes, and H. Leffler. 1993. L-29, a
soluble lactose-binding lectin, is phosphorylated on serine 6 and serine 12 in vivo and by
casein kinase 1. J. Biol. Chem. 268:26712-26718.

Imamoto, N., T. Tachibana, M. Matsubase, and Y. Yoneda. 1995. A karyophilic protein
forms a stable complex with cytoplasmic components prior to nuclear pore binding. J.
Biol. Chem. 270:8559-8565.

Iovine, M. K. and S. R. Wente. 1997. A nuclear export signal in Kap95p is required for
both recycling the import factor and interactions with the nucleoporin GLFG repeat
regions ofNup116p and Nup100p. J. Cell Biol. 137:797-811.

Izaurralde, E., A. Jarmolowski, C. Beisel, I. W. Mattaj, G. Dreyfuss, and U. Fischer.
1997. A role for the M9 transport signal of hnRNP A1 in mRNA nuclear export. J. Cell
Biol. 137:27-35.

Izaurralde, E., J. Lewis, C. Garnberi, A. Jarmolowski, C. McGuigan, and I. W. Mattaj.
1995. A cap-binding protein complex mediating U snRN A export. Nature 376:709-712.

Izaurralde, E., J. Lewis, C. McGuigan, M. Jankowska, E. Darzynkiewicz, and l. W.
Mattaj. 1994. A nucler cap bining protein complex involved in pre-mRNA splicing. Cell
78:657-668.

Izaurralde, E. and I. W. Mattaj. 1995. RNA export. Cell 81:153-159.

Izaurralde, E., J. Stepinski, E. Darzynkiewicz, and I. W. Mattaj. 1992. A cap binding
protein that may mediate nuclear export of RNA polymerase II-transcribed RNAS. J. Cell
Biol. 118:1287-1295.

Jarmolowski, A., W. C. Boelens, E. Izaurralde, and I. W. Mattaj. 1994. Nuclear export of
different classes of RNA is mediated by specific factors. J. Cell Biol. 124:627-635.

Jia, S. and J. L. Wang. 1988. Carbohysrate binding protein 35. Complementary DNA
sequence reveals homology with proteins of the heterogeneous nuclear RNP. J. Biol.
Chem. 263:6009-6011.

120

Jiang, L.-W. and M. Schindler. 1987. Fluorescence photobleaching analysis of nuclear
transport: dynamic evidence for auxiliary channels in detergent-treated nuclei.
Biochemistry 26: 1546-1551 .

Kalirrich, J. F. and M. G. Douglas. 1989. In vitro translocation through the yeast nuclear

envelope: signal-dependent transport requires ATP and calcium. J. Biol. Chem.
264:17979—17989.

Kanigsberg, A., J. Seetharaman, H. Leffler, S. H. Barondes, and J. M. Rini. 1997.
Submitted

Kasai, K.-i. and J. Hirabayashi. 1996. Galectins: A family of animal lectins that decipher
glycocodes. Journal of Biochemistry 119:1-8.

Kayden, C. S., J. Seetharaman, H. Leffler, S. H. Barondes, and J. M. Rini.1997.
Submitted

Kim, F. J., A. A. Beeche, J. J. Hunter, D. J. Chin, and T. J. Hope. 1996. Characterization
of the nuclear export signal of human T-cell lymphotropic virus type 1 Rex reveals that

nuclear export is mediated by position-variable hydrophobic interaction. Mal. Cel. Biol.
16:5147-5155.

Koepp, D. M. and P. A. Silver. 1996. A GTPase controlling nuclear trafficking: running
the right way or walking RANdomly? Cell 87:1-4.

Kraemer, D., R. W. Wozniak, G. Blobel, and A. Radu. 1994. The human CAN protein, a
putative oncogene product associated with myeloid leukemogeneisis, is a nuclear pore
complex protein that faces the cytoplasm. Proc. Natl. Acad. Sci. USA 91 :1519-1523.

Kutay, U., E. Izaurralde, F. R. Bischoff, I. W. Mattaj, and D. GOrlich. 1997. Dominant-
negative mutants of importin-b block multiple pathways of import and export through the
nuclear pore complex. EMBO J. 16:1153-1163.

Laing, J. G. and J. L. Wang. 1988. Identification of carbohydrate binding protein 35 in
heterogeneous nuclear ribonucleoprotein complex. Biochemistry 27:5329—5334.

121

Lee, M. S., M. Henry, and P. A. Silver. 1996. A protein that shuttles between the nucleus
and the cytoplasm is an important mediator of RNA export. Genes & Devel. 10:1233-
1246.

Lefller, H. 1997. Introduction to galectins. Trends in Glycoscience and Glycotechnology
9:9—19.

Lewis, J. D., D. GOrlich, and I. W. Mattaj. 1996. A yeast cap binding protein complex
(yCBC) acts at an early step in pre-mRNA splicing. Nucleic Acids Res 24:3332-3336.

Li, J., G. Elberg, D. C. Crans, and Y. Shechter. 1996. Evidence for the distinct
vanadyl(+4)-dependent activating system for manifesting insulin-like effects.
Biochemistry 35: 8314-8318

Liao, D.-I., G. Kapadia, H. Ahmed, G. R. Vasta, and O. Herzberg. 1994. Structure of S-
lectin, a developmentally regulated vertebrate B-galactoside-binding protein. Proc. Natl.
Acad. Sci. USA 91:1428-1432.

Liao, X. C., H. V. Colot, Y. Wang, and M. Rosbash. 1992. Requirements for U2 snRNP
addition to yeast pre-mRNA. Nucleic Acids Res 20:4237-4235.

Lindstedt, R., G. Apdaca, S. H. Barondes, K. E. Mostov, and H. Leffler. 1993. Apical
secretion of a cytosolic protein by Madin-Darby canine kidney cells. J. Biol. Chem.
268:11750-11757.

Lobsanov, Y. D., M. A. Gitt, H. Leffler, S. H. Barondes, and J. M. Rini.1993. X-ray
crystal structure of the human dimeric S-lac lectin, L-14-II, in complex with lactose at
2.9-A resolution. J. Biol. Chem. 268:27034-27038.

Lobsanov, Y. D. and J. M. Rini. 1997. Galectin structure. Trends in Glycoscience and
Glycotechnology 9: 145-154.

Lotz, M. M., C. W. Andrews,Jr., C. A. Korzelius, E. C. Lee, G. D. Steele,Jr., A. Clarke,
and A. M. Mercurio. 1993. Decreased expression of Mac-2 (carbohydrate binding protein
35) and loss of its nuclear localization are associated with the neoplastic progression of
colon carcinoma. Proc. Natl. Acad. Sci. U. S. A. 90:3466-3470.

122

Macara, I. G., K. Kustin, and J.C. Cantley Jr. 1980. Glutathione reduces cytoplasmic
vanadate. Mechanism and physiological implications. Biochim. Biophys. Acta 629:95-106

Mandel], R. B. and C. M. Feldherr. 1990. Identification of two HSP70-related Xenopus
oocyte proteins that are capable of recycling across the nuclear envelope. J. Cell Biol.
111:1775-1783.

Marshallsay, C. and R. Liihrmann. 1994. In vitro nuclear import of snRNPs: cytosolic
factors mediate m3G-cap dependence of U1 and U2 snRNP transport. EMBO J. 13:222-
231.

Mayeda, A. and A. R. Krainer. 1992. Regulation of alternative pre-mRNA splicing by
hnRNP Al and splicing factor SF2. Cell 68:365-375.

Melchior, F., T. L. Guan, N. Yokoyarna, T. Nishirnoto, and L. Gerace. 1995. GTP
hydrolysis by Ran occurs at the nuclear pore complex in an early step of protein import.
J. Cell Biol. 131:571-581.

Melchior, F., B. M. Paschal, J. Evans, and L. Gerace. 1993. Inhibition of nuclear protein
import by nonhybrolyzable analogues of GTP and identification Of the small GTPase
Ran/TC4 as an essential transport factor. J. Cell Biol. 123:1649-1659.

Meyer, B. E. and M. H. Malirn. 1994. The HIV-1 Rev trans-activator shuttles between the
nucleus and the cytoplasm. Genes-Dev. 8: 1538-1 547.

Michaud, S. and R. Reed. 1991. An ATP-independent complex commits pre-mRNA to
the mammalian spliceosome assembly pathway. Genes & Devel. 5:2534-2546.

Michael, W. M., M. Y. Choi, and G. Dreyfuss. 1995. A nuclear export signal in hnRNP
A1: A signal-mediated, temperature-dependent nuclear protein export pathway. Cell
83:415-422.

Michael, W. M., P. S. Eder, and G. Dreyfuss. 1997. The K nuclear shuffling domain: a
novel signal for nuclear import and nuclear export in the hnRNP K protein. EMBO J.
16:3587-3598.

Misaki, Y., G. J. Pruijn, A. W. van der Kemp, and W. J. van Venrooij. 1994. The 56K
autoantigen is identical to human annexin XI. J. Biol. Chem. 269:4240-4246.

123

Mizutani, A., N. Watanabe, T. Kitao, H. Tokunritsu, and H. Hidaka. 1995. The long
amino-terminal tail domain of annexin XI is necessary for its nuclear localization. Arch.
Biochem. Biophys. 318: 1 57-165.

Moore, M. J., C. C. Query, and P. A. Sharp. 1993. Splicing of precursors of mRNA by
the spliceosome. In The RNA World. R. F. Gesteland and J. F. Atkins, editors. Cold
Spring Harbor Laboratory Press, New York. 303-357.

Moore, M. S. 1996. Protein translocation: Nuclear export - out of the dark. Current
Biology 6: 137-140.

Moore, M. S. and G. Blobel. 1993. The GTP-binding protein Ran/T C4 is required for
protein import into the nucleus. Nature 365:661-663.

Moore, M. S. and G. Blobel. 1994. Purification of a Ran-interacting protein that is
required for protein import into the nucleus. Proc. Natl. Acad. Sci. USA 91 : 10212-10216.

Moroianu, J. and G. Blobel. 1995. Protein export from the nucleus requires the GTPase
Ran and GTP hydrolysis. Proc. Natl. Acad Sci. USA 92:4318-4322.

Moroianu, J ., G. Blobel, and A. Radu. 1995. Previously identified protein of uncertain
function is karyopherin a and together with karyopherin b docks import substrate at
nuclear pore complexes. Proc. Natl. Acad. Sci. USA 92:2008-2011.

Moroianu, J ., G. Blobel, and A. Radu. 1996. The binding site of karyopherin a for

karyopherin b overlaps with a nuclear localization sequence. Proc. Natl. Acad. Sci. USA
93:6572-6576.

Moroianu, J ., G. Blobel, and A. Radu. 1996. Nuclear protein import: Ran-GTP dissociate
the karyopherin ab heterodimer by displaing a from and overlapping binding site on b.
Proc. Natl. Acad. Sci. USA 93:7059-7062.

Moroianu, J., M. Hijikata, G. Blobel, and A. Radu. 1995. Mammalian karyopherin a,b
and ab heterodimers: al or a2 sububit binds nuclear localization signal and b subunit

interacts with peptide repeat-containing nucleoporins. Proc. Natl. Acad Sci. USA
92:6532-6536.

124

Mostov, K., G. Apodaca, B. Aroeti, and C. Okarnoto. 1992. Plasma membrane protein
sorting in polarized epithelial cells. J. Cell Biol. 116:577-583.

Moutsatsos, I. K., J. M. Davis, and J. L. Wang. 1986. Endogenous lectins from cultured
cells: Subcellular localization of carbohydrate-binding protein 35 in 3T3 fibroblasts. J.
Cell Biol. 102:477-483. '

Moutsatsos, I. K., M. Wade, M. Schindler, and J. L. Wang. 1987. Endogenous lectins
from cultured cells: nuclear localization of carbohydrate-binding protein 35 in
proliferating cells. Proc. Natl. Acad. Sci. USA 84:6452-6456.

Muesch, A., E. Hartrnann, K. Rohde, A. Rubartelli, R. Sitia, and T. A. Rapoport. 1990. A
novel pathway for secretory proteins? TIBS 15:86-88.

Nehrbass, U. and G. Blobel. 1996. Role of the nuclear transport factor p10 in nuclear
import. Science 272: 120-122.

Neuman de Vegvar, H. E. and J. E. Dahlberg. 1990. Nucleocytoplasmic transport and
processing of small nuclear RNA precursors. Mol. Cel. Biol. 10:2265-3375.

Newmeyer, D. D. and D. J. Forbes. 1988. Nuclear import can be separated into distinct
steps in vitro: nuclear pore binding and translocation. Cell 52:641-653.

Nigg, E. A. 1997. Nucleocytoplasmic transport: Signals, mechanisms and regulation.
Nature 386:779-787.

Numinskaya, M. and T. F. Linsenmayer. 1996. Identification of characterization of up-
regulated genes during Chondrocyte hypertrophy. Developmental Dynamics 206:260-271.

O'Farrell, P. Z., H. M. Goodman, and P. H. O’Farrell. 1977. High resolution two-
dimensional electrophoresis of basic as well as acidic proteins. Cell 1977 12:1133-1141

Pante, N., R. Bastos, I. McMorrow, B. Burke, and U. Aebi. 1994. Interactions and three

dimensional localization of a group of nuclear pore complex proteins. J. Cell Biol.
126:603-617.

125

Paschal, B. M. and L. Gerace. 1995. Identification of NTF2, a cytosolic factor for nuclear
import that interacts with nuclear pore complex protein p62. J. Cell Biol. 129:925-93 7.

Patterson, R. J ., S. F. Dagher, A. Vyakamam, and J. L. Wang. 1997. Nuclear galectins:
Functionally redundant components in processing of pre-mRN A. Trends in Glycoscience
and Glycotechnology 9:77-85.

Perillo, N. L., K. E. Pace, J. J. Seilharner, and L. G. Batun. 1995. Apoptosis of T cells
mediated by galectin-l. Nature 378:736-739.

Perillo, N. L., C. H. Uittenbogaart, J. T. Nguyen, and L. G. Baum. 1997. Galectin-1, an
endogenous lectin produced by thymic epithelial cells, induces apoptosis of human
thymocytes. J. Exp. Med. 185:1851-1858.

Peters, R 1986. Fluorescence rnicrophotolysis to measure nucleocytoplasrrric transport
and intracellular mobility. Biochim. Biophys. Acta 864:305-359.

Pifiol-Roma, S., Y. D. Choi, M. J. Matunis, and G. Dreyfuss. 1988. Immunopurification
of heterogeneous nuclear ribonucleoprotein particles reveals an assortment of RNA-
binding proteins. Genes-Dev. 2:215-227.

PiflOl-Roma, S. and G. Dreyfuss. 1993. Shuttling of pro-mRNA binding proteins between
nucleus and cytoplasm. Nature 355:730-732.

Piotrowicz, R. S., J. L. Martin, W. H. Dilhnan, and E. G. Levin. 1997. The 27-kDa heat
shock protein facilitates basic fibroblast growth factor release from endotlrelial cells. J.
Biol. Chem. 272:7042-7047.

Pollard, V. W., W. M. Michael, S. NAkielny, M. C. Sionri, F. Wang, and G. Dreyfuss.
1996. A novel receptor-mediated nuclear import pathway. Cell 86:985-994.

Powers, M. A., D. J. Forbes, J. E. Dalrlberg, and E. Lund. 1997. The vertebrate GLFG
nucleoporin, Nup98, is an essential component of multiple RNA export pathway. J. Cell
Biol. 136:241-250.

Powers, M. A., C. Macaulay, F. R. Masiarz, and D. J. Forbes. 1995. Reconstituted nuclei
depleted of a vertebrate GLF G nuclear pore protein, p97, import but are defective in
nuclear growth and replication. J. Cell Biol. 128:721-736.

126

Puch, A. C. and B. Key. 1995. Identification of cells expressing galectin-l, a galactose-
binding receptor, in the rat Olfactory system. The Journal of Comparative Neurology
357:513-523.

Puche, A. C., F . Poirier, M. Hair, P. F. Bartlee, and B. Key. 1996. Role of galectin-1 in
the developing mouse olfactory system. Developmental Biology 179:274-287.

Radu, A., G. Blobel, and R. W. Wozniak. 1993. NuplSS is a novel nuclear pore complex
protein that contains neither repititive sequence motifs nore reacts with WGA. J. Cell
Biol. 121:1-9.

Radu, A., G. Blobel, and R. W. Wozniak. 1994. Nup107 is a novel nuclear pore complex
protein that contains a leucine zipper. J. Biol. Chem. 269:17600-17605.

Radu, A., M. S. Moore, and G. Blobel. 1995. The peptide repeat domain of nucleoporin

Nup98 functions as a docking site in transport across the nuclear pore complex. Cell
81:215-222.

Rexach, M. and G. Blobel. 1995. Protein import into nuclei: association and dissociation
reactions involving transport substrate, transport factors, and nucleoporins. Cell 83:683-
692.

Richards, S. A., K. L. Carey, and I. G. Macara. 1997. Requirement of guanosine
triphosphate-bound Ran for signal-mediated nuclear protein export. Science 276: 1842-
1844.

Richardson, W. D., A. D. Mills, S. M. Dilworth, R. A. Laskey, and C. Dingwall. 1988.
Nuclear protein migration involves two steps: rapid binding at the nuclear envelope
followed by slower translocation through nuclear pores. Cell 52:655-664.

Roff, C. F ., P. R. Rosevear, J. L. Wang, and R. Barker. 1983. Identification of
carbohydrate-binding proteins from mouse and human fibroblasts. Biochem. J. 211:625-
629.

Rothmarr, J. E. and L. Orci. 1992. Molecular dissection of the secretory pathway. Nature
355:409-415.

127

Sachs, A. B., P. Samow, and M. W. Hentze. 1997. Starting at the beginning, middle, and
end: translation initiation in eukaryocytes. Cell 89:831-83 8.

Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning. Cold Spring
Harbor Laboratory Press, New York.

Siders, W. M. and S. B. Mizel. 1995. Interleukin-1b secretion. J. Biol. Chem. 270:16258-
16264.

Sirnons, K. and A. Wandinger-Ness. 1990. Polarized sorting in epithelia. Cell 62:207-
210.

Sionri, H. and G. Dreyfuss. 1995. A nuclear localization domain in the hnRNP A1
protein. J. Cell Biol. 129:551-560.

Smalheiser, N. R. 1996. Proteins in unexpected locations. Mol. Biol. Cell 7:1003-1014.

Snow, C. M., A. Senior, and L. Gerace. 1987. Monoclonal antibodies identify a group Of
nuclear pore complex glycoproteins. J. Cell Biol. 104:1143-1156.

Spector, D. L. 1996. Nuclear organization and gene expression. Exp. Cell Res. 229:189-
197.

Starr, C. M., M. D'Onofrio, M. K. Park, and J. A. Hanover. 1990. Primary sequence and
heterologous expression of nuclear pore glycoprotein p62. J. Cell Biol. 1 10: 1861-1871.

Sukegawa, J. and G. Blobel. 1993. A nuclear pore complex protein that contains zinc
fingers and faces the nucleoplasm. Cell 72:29-38.

Tarantini, F., S. Gamble, A. Jackson, and T. Maciag. 1995. The cysteine residue
responsible for the release of fibroblast growth factor-1 resides in a domain independent
of the domain for phsophatidylserine binding. .1. Biol. Chem. 270:29039-29042.

Tokunritsu, H., A. Mizutarri, H. Minarni, R. Kobayashi, and H. Hidaka. 1992. A
calcyclin-associated protein is a newly identified member of the Cazi/phospholipid-
binding proteins, annexin family. J. Biol. Chem. 267:8919-8924.

128

Valcarcel, J. and M. R. Green. 1996. The SR protein family: pleiotropic functions in pre-
mRNA splicing. TIBS 21:296-301.

Visa, N., A. Alzhanova-Ericsson, X. Sun, E. Kiseleva, B. BjOrkroth, T. Wurtz, and B.
Daneholt. 1996. A pre-mRNA-binding protein accompanies the RNA from the gene
through the nuclear pores and into polysomes. Cell 84:253-264.

Visa, N., E. Izaurralde, J. Ferreira, B. Daneholt, and I. W. Mattaj. 1996. A nuclear cap-
binding complex binds Balbiani ring pre-mRNA cotranscriptionally and accompanies the
ribonucleoprotein particle during nuclear export. J. Cell Biol. 133:5-14.

Vyakamam, A. 1996. Galectin-l in the cell nucleus: Evidence for a role in pre-mRNA .
splicing. Ph. D. Dissertation.

Vyakamam, A., S. F. Dagher, J. L. Wang, and R. J. Patterson. 1997. Evidence for a role
for galectin-l in pre-mRNA splicing. Mal. Cel. Biol. 17:4730-4737.

Wada, J., K. Ota, A. Kumar, E. I. Wallner, and Y. S. Kanwar. 1997. Developmental
regulation, expression, and apoptotic potential of galectin-9, a B-galactoside binding
lectin. J. Clin Invest. 99:2452-2461.

Wang, J. L., Anandita, S.-Y. Wang, and N. Agrwal. 1993. Animal lectins: Carbohydrate
binding protein 35: Production of the recombinant polypeptide and its individual
domains. In Lectins: Biology, biochemistry, clinical biochemistry. J. Basu, M. Kundu,
and P. Chakrabarti, editors. Wiley Eastern Limited, New Delhi. 89-99.

Wang, J. L., J. G. Laing, and R. L. Anderson. 1991. Lectins in the cell nucleus.
Glycobiology. 1:243-252.

Wang, L., H. Inohara, K. J. Pienta, and A. Raz. 1995b. Galectin-3 is a nuclear matrix
protein which binds RNA. Biochem. Biophys. Res. Commun. 217:292-303.

Wang, S.-Y., P. G. Voss, R. J. Patterson, and J. L. Wang. 1995a. Studies on the cell
surface versus nuclear localization of galectin-3. Antibody, Immunoconjugates, and
Radiopharmaceuticals 8 :31 1-3 24.

129

Weighardt, F., G. Biarnonti, and S. Riva. 1996. The roles of heterogeneous nuclear
ribonucleoproteins (hnRNP) in RNA metabolism. BioEssays 18:747-756.

Weis, K., C. Dingwall, and A. I. Lamond. 1996. Characterization of the nuclear protein

import mechanism using Ran mutants with altered nucleotide binding specificities.
EMBOJ. 15:7120—7128.

Weller, P. F., D. S. Bach, and K. F. Austen. 1984. Biochemical characterization of hmnan

eosinophil Charcot-Leyden crystal protein (lysophospholipase). J. Biol. Chem.
259:15100-15105.

Wen, W., J. L. Meinkotlr, R. Y. Tsien, and S. S. Taylor. 1995. Identification of a signal
for rapid export of proteins from the nucleus. Cell 82:463-473.

Wilken, N., U. Kossner, J.-L. Senécal, U. Scheer, and M.-C. Dabauvalle. 1993. Nup180,
a novel nuclear pore complex protein localizing to the cytoplasmic ring and associated
fibrils. J. Cell Biol. 123:1345-1354.

Wolff, B., M. C. Willingharn, and J. A. Hanover. 1988. Nuclear protein import:
specificity for transport across the nuclear pore. Exp. Cell Res. 178:3 1 8-3 34.

Wozniak, R. W., E. Bartrrik, and G. Blobel. 1989. Primary structure analysis of an
integral membrane glycoprotein of the nuclear pore. J. Cell Biol. 108:2083-2092.

Wozniak, R. W. and G. Blobel. 1992. The single transmembrane segment of ngIO is

sufficient for sorting to the pore membrane domain of the nuclear envelope. J. Cell Biol.
119:1441-1449.

Wu, J., M. J. Matunis, D. Kraemer, G. Blobel, and E. Coutavas. 1995. Nup358, a
cytoplasmc exposed nucleoporin with peptide repeats, Ran-GTP binding sites, zinc
fingers, a cyclophilin A homologous domain, and a leucine-rich region. J. Biol. Chem.
270:14209-14213.

Wu, J. Y. and T. Maniatis. 1993. Specific interactions between proteins implicated in
splice site selection and regulated altemative splicing. Cell 75: 1061-1070.

Yang, J., J. Liu, and D. B. DeFranco. 1997. Subnuclear trafficking of glucocorticoid
receptors in vitro: Chromatin recycling and nuclear export. J. Cell Biol. 137:523-538.

130

Yang, R.-Y., D. K. Hsu, and F.-T. Liu. 1996. Expression of galectin-3 modulates T-cell
growth and apoptosis. Proc. Natl. Acad. Sci. USA 93:6737-6742.

Yaseen, N. R. and G. Blobel. 1997. Cloning and characterization of human karyopherin
b3. Proc. Natl. Acad. Sci. USA 94:4451-4456.

Yin, X.-M., Z. N. Oltval, and S. J. Korsmeyer. 1994. 8H] and BH2 domains of Bcl-2 are
required for inhibiton of apoptosis and heterodirnerization with Bax. Nature 369:321-323.

Yokoyama, N., N. Hayashi, T. Seki, N. Pante, T. Ohba, K. Nishii, K. Kuma, T.
Hayashida, T. Miyata, U. Aebi, M. Fukui, and T. Nishirnoto. 1995. A giant nucleopore
protein that binds Ran/TC4. Nature 376: 184-1 88.

Yoneda, Y., N. Imamato-Sonobebe, Y. Matsuoka, R. Iwarnoto, YiKiho, and T. Uchida.
1988. Antibodies to Asp-Asp-Glu-Asp can inhibit transport of nuclear proteins into the
nucleus. Science 242:276-278.

Zilhnann, M., M. L. Zapp, and S. M. Berget. 1988. Gel electrophoretic isolation of
splicing complexes containing U1 small nuclear ribonucleoprotein particles. Mal. Cel.
Biol. 8:814-821.

Zolotukhin, A. S. and B. K. Felber. 1997. Mutations in the nuclear export signal of
human Ran-binding protein RanBPl block the Rev-mediated posttranscriptional
regulation of human immunodeficiency virus type 1. J. Biol. Chem. 272:11356-11360.

MICHIGAN srnTE UNIV. LIBRARIES
lllllllllllllllllllllllllllllllllllllllllllllllllllllllll
31293017126511